Wikipedia talk:No original research

Archives

See also Wikipedia talk:Proposal to replace No Original Research
For the "Sandbox", click here.

Verifiability, not truth??

This is quoted like a mantra, but Wikipedia DEFINES "verifiable" in terms that have no reference except to objective truth. For example: it DEFINES verifiable as "reliable" and from "established experts" and from sources "with a reputation for accuracy". Well, the words "reliable" and "expert" and "accurate" have no meaning unless we're talking about truth-relatedness. Truthiness, as Colbert would say. So what's with the "verifiable, NOT true" crap? Is somebody just trying to confuse things? There's no room for a NOT in that sentence, if the said definition of verifiable is to be used. Would somebody like to explain what point is trying to be made in this statement, at such labor? Cause I'm obviously not getting it. SBHarris 05:51, 26 November 2007 (UTC)

I think the point is that if editors try to be the arbiter of truth themselves, that would be engaging in OR. We must look for consensus that has formed organically in the RS, and not try to judge truth for ourselves. Some common sense is required in applying this, however. Out-of-date sources may have verifiable information that is just plain wrong for a number of reasons, and we can ignore this in favor of more up-to-date material. Sources sometimes make obvious errors, and we can ignore these when we have sources that did not make or repeat the mistake. So if something is verifiable but not true, there should be RS to back that assessment. You can't substitute your own unpublished assessment, however. Dhaluza (talk) 12:57, 27 November 2007 (UTC)

I believe the operative word here is "reputation" for accuracy, which is a different thing from accuracy itself. The distinction is critical. Reputations are with respect to communities and are not universal. A doctor of theology certified by the Vatican and peer-reviewed by other theologians may have a reputation for reliably reflecting Catholic doctrine, but regarding such an individual as a reliable source on Catholic doctrine in no way implies that Wikipedia believes Catholic doctrine is "the truth." Similarly, a PhD peer-reviewed by the academic community as an expert on biology can reliably reflect scientific views on that subject, but this doesn't mean Wikipedia endorses everything biologists say either. The issue is simply whether the communities in which the reputations are formed represent points of view which are "significant" as WP:NPOV defines that term. WP:NPOV says more than simply that editors can't decide who to believe, it says that Wikipedia shouldn't decide this. The policy is to present diametrically opposed points of view so long as they are significant; Creationism and Evolution; Libertarianism and Natural law theory; Monotheism, Polytheism, and Atheism; Capitalism and Socialism; Materialism and Idealism; Rationalism and Mysticism; you name it. The encyclopedia isn't interested in who is right in the many debates about what the truth is and how best to perceive it. It simply presents the different points of view. Best, --Shirahadasha 23:09, 2 December 2007 (UTC)

Let me see if I got this straight...

It's all a semantics problem with respect to the definition of the words "primary" and "secondary" sources?

What types of sources are acceptable?

What types of sources are unacceptable?

Please be specific.

And...

What exactly is "original research"?

The Transhumanist    05:57, 27 November 2007 (UTC)

It's not simply a semantics problem, it's also a syntactics problem--i.e. where is the dividing line between primary and secondary, and is it a clear line, or is there a gray area between? The answer to your questions is that all reliable sources are acceptable, but they must be used appropriately. As far as what is original research, that is WP content whose source is the contributing editor, not another published author. It really should have nothing to do with the proper use of sources, because if there is another published source, it's not original to the editor. Dhaluza (talk) 12:42, 27 November 2007 (UTC)

It's true that the dividing line is not clearly drawn. Similarly, the line between original prose and original research is not clearly drawn, nor is the line between NPOV and POV, nor between due weight and undue weight, nor between reliable and unreliable sources, nor for that matter the line between verifiable and verified with an inline citation (i.e., what kind of material must be cited in response to a "citation-needed" template vs. removing the template with words to the effect of "gimme a break-- no citation is needed for the reference to Ottawa as the capital city of Canada"). All these are negotiated as necessary among participants in an article. ... Kenosis (talk) 15:58, 27 November 2007 (UTC)

It's also true that primary sources can be particularly open to synthesis by selective quotation, an example being racism where a series of quotations from 19th century authors are presented in a way that implies that their statements are "academic racism". A secondary source is needed to support this presentation, which in at least one instance is inaccurate. .. dave souza, talk 16:36, 27 November 2007 (UTC)

And how does using primary sources in this context differ from doing the same thing with only secondary sources? The same point could still be implied with selective quotations based solely on secondary sources. Why are primary sources more problemmatic? wbfergus ^Talk 17:06, 27 November 2007 (UTC)

Because citing primary sources exclusively forces one almost always to violate NOR. That is why caution is needed as the potential for abuse (i.e. a novel synthesis to advance a position) is greater. ≈ jossi ≈ (talk) 17:24, 27 November 2007 (UTC)

Maybe jossi is right for the way he defines primary source, but I think it's quite possible to write an article that depends only on what could be considered primary sources (by some definitions) without any original research. Gerry Ashton (talk) 17:44, 27 November 2007 (UTC)

Yes of course. It's totally permissible so long as the statements derived from primary sources "only make descriptive claims about the information found in the primary source, the accuracy and relevance of which is easily verifiable by any reasonable, educated person without specialist knowledge," "Where interpretive claims, analysis, or synthetic claims about primary sources are included in Wikipedia articles, use secondary [or tertiary] sources rather than original analysis by Wikipedia editors."... Kenosis (talk) 17:51, 27 November 2007 (UTC)

Interpretive claims about any source need to be sourced. Examples include "this author is wrong" or "this author was influenced by that other author.". Interpretive claims about the subject matter of the article, on the other hand, can only be found in secondary sources (by definition), and in practice are just cited directly to those sources without requiring additional secondary sources. — Carl (CBM · talk) 17:57, 27 November 2007 (UTC)

CBM wrote "Interpretive claims about the subject matter of the article, on the other hand, can only be found in secondary sources (by definition)...." Not always. If the source is reliable it may be used. If the source made an interpretive claim about the subject matter of the article, and made that claim based on the direct observation of phenomena, rather than reading primary sources, the source is a primary source. --Gerry Ashton (talk) 18:04, 27 November 2007 (UTC)

Gerry wrote: "If the source made an interpretive claim about the subject matter of the article, and made that claim based on the direct observation of phenomena, rather than reading primary sources, the source is a primary source."

Actually,... the raw observation is primary. The analysis/interpretation is secondary.

But such a distinction is irrelevant with respect to OR.

• "X observed^[p] that the typeface used by Wikipedia is sans-serif, from which he concluded^[s] that there wasn't enough lead for the serifs."
  ^^^ both primary AND secondary AND valid
• "[X observed that] the typeface used by Wikipedia is sans-serif.^[p]"
  ^^^ only primary but still valid
• "[X concluded that] Wikipedia doesn't have enough lead for serif fontfaces.^[s]"
  ^^^ only secondary but still valid
• "[X concluded that] serif typefaces require more lead than sans-serif typefaces.^[s]"
  ^^^ OR. Irrelevant whether source is "primary" or "secondary."
• "citing primary sources exclusively forces one almost always to violate NOR"
  ^^^ OR *and* non-V. Irrelevant whether (non-existent) source is "primary" or "secondary."

See? Not exactly rocket science, right? -- Fullstop (talk) 20:00, 27 November 2007 (UTC)

1. All sources that meet WP:RS are acceptable. The 'S' in 'RS' stands for "sources."
2. "Original research" is the creative "ingredient" that, when added to 'x', results in 'x±n'.
   The result then says something that is not explicitly evident in the source.

If you will, from the point of view of attribution, "original research" is plagiarism in reverse: Where 'plagiarism' is the attribution of someone else's words/ideas to one's self, 'original research' is the attribution of one's own words/ideas to someone else.

Original research is very easy to spot.

-- Fullstop (talk) 17:55, 27 November 2007 (UTC)

All sources that meet WP:RS are not equal. Answers in Genesis is a RS for an article or section about themselves, but a secondary source is needed to give the third party view required for WP:V and to avoid original research in evaluating their claims. .. dave souza, talk 19:14, 27 November 2007 (UTC)

That's because Answers in Genesis is questionably reliable source about themselves. It has nothing to do with original research, or whether they are a primary or secondary source. I wouldn't trust Answers as a secondary source about themselves, either. For example, if their website quoted one of their pamphlets, or even a newspaper blurb about them, I wouldn't trust it any more than if they presented the same information in a more primary way. I would trust the newspaper cited as a primary source about Answers, just not as a source filtered secondarily through Answers. COGDEN 18:46, 29 November 2007 (UTC)

I don't wish to give overly short shrift to assertions that, logically speaking, PSTS isn't strictly necessary or should be a guideline for allocating sources in complex, obscure and/or controversial articles. But, logically speaking, most of the content of all of the three core content policies isn't absolutely needed in order to perform the function of cooperatively writing an online encyclopedia that's neutral with "due weight", all the content of which is verified or at least verifiable from reliable sources. Logically, there's no need for WP:NOR at all. It's already covered under verifiability and reliable sources! Sweet, no?.

Then we can get right to work defining what is a reliable source wiki-wide, because reliability is a totally arbitrary and arguable concept. Perhaps it should be eliminated too, on the grounds of being arbitrary and arguable. After that, I think maybe WP:V should be eliminated too. Why? Because the four words "stick to the sources" along with additional three words "no personal opinions" (seven words total) are so parsimonious that they can readily be appended to WP:NPOV. And there's no need even for that much, because if the point is using a "NPOV without undue weight to minority positions, using only published reliable sources", it should be obvious that personal opinions, personal speculations and personal analyses can't be used. So we can cut all that verbiage down to "Be neutral and objective, and stick to the reliable sources". No need for WP:NOR, no need for WP:V, only for a very short WP:Editorial policy and a lengthy WP:RS analysis to account for the wide range of reiiability criteria that folks may have. So, most of the debate really comes down to "What is a reliabile source?" No?

Believe it or not, I'm only half-jesting. Editorial policies do not always lend themselves to strict logical analyses. It's more of a balancing act, especially across something as large and broad as the wiki. And on top of it, heck, maybe the three core policies should all be called guidelines. Policies sounds too strict for things that are this hard to pin down logically.

Pardon my sarcasm please. But these are editorial policies and they don't lend well to strict logical analyses. ... Kenosis (talk) 19:50, 27 November 2007 (UTC)

This talk page has gone to the dogs somewhat. Pretty much impossible to see who's saying what. :-( SlimVirgin ^{(talk)(contribs)} 20:23, 27 November 2007 (UTC)

Of course it has. The other symptom is the failure to check the context of whatever it is that the poor beasts think they are responding to. -- 22:07, 27 November 2007 (UTC) —Preceding unsigned comment added by Fullstop (talk • contribs)

Kenosis, your point makes some sense. It is comparable to known problems in social philosophy and political science (which are relevant to understanding Wikipedia). The most logical solution to a problem is rarely the pragmatic solution. The most practical solution is rarely the socially acceptable solution. "Acceptable" solutions are rarely pleasing to more than a small minority, but usually tolerable and/or accommodating to most system participants. Vassyana (talk) 20:54, 27 November 2007 (UTC)

Thanks for that. I'm tempted to want to put a {{citation needed}} after "It is comparable to known problems in social philosophy and political science (which are relevant to understanding Wikipedia)" ;-) I agree it's comparable. Thank you for that thoughtful insight. ... Kenosis (talk) 21:20, 27 November 2007 (UTC)

Responding to Kenosis above, logically, NOR is not entirely covered by WP:V, and there are very good reasons for keeping it that have nothing to do with PSTS. Here's why we need both policies: while anything that is original research is necessarily unverifiable, the opposite is not always true. WP:V focuses on making sure every fact can be verified outside of Wikipedia. One thing that cannot be verified outside Wikipedia is original research, but that's not the only thing. PSTS is in the realm of "other things" that are arguably not "verifiable" (i.e. "reliable") but are also not original research, because if something is stated in a primary or secondary source, it's obviously not original. The only thing keeping PSTS in OR, and not V, is the unproven and dubious theory that statistically, citation to primary sources is more likely to lead to original research than citation to secondary sources. The other dubious argument in favor of PSTS—the theory that primary sources are purportedly less "reliable" than secondary sources, belongs squarely in WP:V. COGDEN 19:07, 29 November 2007 (UTC)

Can someone show me a real example of good primary-source material that could not be used under this policy?

I asked this above, but I'm creating a new section in case it gets lost.

Seems to me that all this theoretical discussion is confusing people, and not helping the policy. I'm guessing we could easily all have what we want with a few tweaks (provided we all understand what the basic terms mean as used by scholars and other researchers, which is how we're using them too).

So could somehow please produce an actual example of where this policy, if adhered to, would have prevented the use of a good source — or did prevent it — so we can see what we're dealing with? SlimVirgin ^{(talk)(contribs)} 22:48, 27 November 2007 (UTC)

I think that's a good idea. Policy is usually bad policy unless it grows organically from real cases. I like to take examples from featured articles, since they can be expected to generally reflect the best of actual, widely-accepted Wikipedia practice. You never have to go very far to find good PSTS violations in featured articles. In today's featured article, winter service vehicle, the fourth sentence states:

“ Airports use winter service vehicles to keep runways and taxiways free of snow and ice, which, besides endangering aircraft takeoff and landing, can interfere with communication equipment. ”

The citation for this sentence is a primary source press release from the Dresden Airport which gives what could be interpreted as an example of such an interference. Since the citation is a primary source, the above sentence is not "merely descriptive", and is also an "interpretation" (a double violation of PSTS), it would be prohibited by the present policy, even though I think everybody would agree that this is a perfectly acceptable citation and this is not original research, since it meets the requirements of WP:V. Anything that meets WP:V cannot be original research. COGDEN 23:37, 27 November 2007 (UTC)

We allow the use of primary sources where there is no analysis needed and the source can be understood by reasonable laypersons. I don't see why your example would violate the current policy. Crum375 (talk) 23:44, 27 November 2007 (UTC)

Cogden, in what way is the use of that primary source a violation of the policy? There is no interpretation of what the source says. The source says X. We say X. SlimVirgin ^{(talk)(contribs)} 00:02, 28 November 2007 (UTC)

I don't think this is a good use of the source, because the Wikipedia article claims that snow and ice can interfere with communication equipment, but the source does not say that. --Gerry Ashton (talk) 00:04, 28 November 2007 (UTC)

This would have nothing to do with primary vs. secondary — if the source is deficient, you need another source or you remove the claim. Crum375 (talk) 00:11, 28 November 2007 (UTC)

Exactly. Non-adherence to source, aka OR, never has anything to do with primary vs. secondary. -- Fullstop (talk) 00:54, 28 November 2007 (UTC)

(edit conflict) Oh, I see, you mean it's a source giving one example, rather than a source talking about airports in general. Yes, if the issue were very contentious, that would be a violation of this policy. If, for example, I wrote: "Scholars do not regard the murder of the Roma as part of the Holocaust," and I linked to one scholar who said "I do not regard the murder of the Roma as part of the Holocaust," that would be a misuse of the source material (whether secondary or primary). In a non-contentious issue, however, it's quite acceptable to cite an example of the kind of thing you're generalizing from, so long as you're sure your generalization is reasonable. SlimVirgin ^{(talk)(contribs)} 00:08, 28 November 2007 (UTC)

The reason the article violates PSTS is that unlike WP:V, the current version of PSTS does not make an exception for obvious and non-controversial interpretations of primary sources. If the claim is not descriptive or it is interpretive (here, it's both), then a primary source is insufficient. Period. Even if the claim is verifiable, which it is here, since the issue is non-contentious. COGDEN 19:13, 29 November 2007 (UTC)

That FA (like far too many) is far from the "best practices" goal. On the contrary, it has some good examples of NOR violations. See Talk:Winter service vehicle#Recommend a source audit. Vassyana (talk) 11:26, 28 November 2007 (UTC)

Oh, so, like, the symptoms should be "cured," rather than the sickness itself? -- Fullstop (talk) 15:20, 28 November 2007 (UTC)

Further, the implied suggestion that FA reviewers are incompetent is a consequence of the supposition that the NOR page is fine, and that the fault must then lie with the reviewers.

-- Fullstop (talk) 15:20, 28 November 2007 (UTC)

What are you considering to be the "symptoms" and "sickness"? I agree with Vassyana that the Winter service vehicle article contains some examples of text that is likely original research. It's not really important to assign blame for OR existing there, but the problem should be addressed. Are you suggesting that the way to deal with original research in that article is to change the NOR policy? Chaz ^Beckett 17:19, 28 November 2007 (UTC)

The point is: what exactly is the purpose of telling us here that that article is busted?

Its rather plain, I think, that the article should not have achieved FA in the first place, but did because neither the editors of the article, nor the FA reviewers had an adequate understanding of OR to be able to avoid/spot it.

Of course "the problem should be addressed," the question is only which problem is relevant here: The problem that allowed OR to happen (the sickness), or the fact that the OR happened (the symptom)?

-- Fullstop (talk) 19:57, 28 November 2007 (UTC)

The point is that it was raised as an example, and upon even cursory examination the use of the source was deeply flawed. Additionally, this is the NOR talk page and the article has/had some serious OR problems. Obviously, the people here have put a lot of thought into what constitutes OR, so it's not out of place to have people who are active here take a look. Finally, the failures of that article have little to nothing to do with the phrasing and presentation of this policy. To be utterly blunt, the references simply weren't checked by the FA reviewers. If you want to fix the illness, this isn't the place to look. Vassyana (talk) 07:59, 29 November 2007 (UTC)

Another example is what I also mentioned above, in your (SlimVirgin's) orginal request. The article United States physiographic region can be greatly expanded, and several new articles created on the physiographic regions of the rest of the world, based almost entirely on primary source material. There is extremely little in the way of secondary material on this subject, as this subject consists almost entirely of maps created by Edwin Raisz, A. Lobeck, and Guy Harold-Smith. I do not know the copywrite status of the maps (one I looked at yesterday was dated 1960, another was dated in the 1930's), so I don't know if I could scan them (in multiple sections, then pasted back together) or not. In order to create any text for an article, I would need to engage in source-based research (of those maps) to compile a list of the various physiographic provinces per continent. I would then need to engage in analysis, synthesis, and explanations to express what is contained on those maps (in the tables along the margins). Once that was accomplished, I could finally start to include some information on some of the indivudual physiographic regions gleaned from secondary sources. However, the 'master' list of all of the various physiographic regions only exists on the various maps (with the exception of North America, which still needs expansion). wbfergus ^Talk 12:19, 28 November 2007 (UTC)

Assuming you resolve the copyright issues, it still sounds like WP:OR to me, which interpreting and analyzing primary material usually is. If you were to publish that in some peer reviewed scientific journal, then you could link to it, but Wikipedia is not intended for original research. Crum375 (talk) 15:49, 28 November 2007 (UTC)

Well, the interpreting, etc. comes from looking at the maps and seeing which provinces and/or sections are contained in which physiographic division. In a few cases, the 'areas' on the map weren't broken out in the table, so one actually had to see the map and see which ones were included where. It's also impossible to describe most of the areas without supplying the map (which are probably copyrighted), so again, in order to describe the boundaries in many cases would violate the OR policy, since the descriptions would be coming from primary materials. Basically, the maps are primary materials, and if they can't be uploaded because of copyrights, then decsriptions should be included. In some cases, as Vassyana says below, I was able to get descriptions of some areas from other sources, but not all areas by a longshot. Even saying "XYZ physiographic section is bordered by these 6 other physiographic 'areas'" would be a violation of this OR policy, since again, it is based on a primary source (the map). However, in the one or two (maybe even three) cases I was able to find, I could probably cite a secondary source that said that, whether the source actually sid it or not, as then I would be allowed to make such judgement calls without the extra restrictions placed upon primary sources. wbfergus ^Talk 18:46, 28 November 2007 (UTC)

wbfergus, physiography is a synonym for physical geography. Under both names, there are plentiful references available, just counting textbooks. A small amount of web searching can yield dozens of sources just by browsing post-secondary course syllabuses. I also agree with Crum375 that the process you describe sounds exactly like original research. Vassyana (talk) 17:56, 28 November 2007 (UTC)

Physiography is much more than merely physical geography. It also encompasses geomorphology and ecosystems as well, besides other components. But, I have tried in the past to find any source that provides a complete and comprehensive list of physiographic regions and their respective components. I have been unable to find one. I was able to find several sites and other sources that talked about a small subset in each one, but no 'master' listing anywhere. And it also took a lot of researching just to find the ones that I did. I have no interest in getting myself published. There's nothing in it for me. I merely would like to make the information available for others, so that they could have a far simpler time of trying to organize the data than I had, and then if they wanted, correct it, add to it, etc., or even publish it themselves. What I have is pretty bare-bones, but by far the most comprehensive list of all sources and the pertinent information in one place that I've found in around 8 years or so. wbfergus ^Talk 18:32, 28 November 2007 (UTC)

Physiography is indeed synonymous with physical geography. Physical geography has the same general definition as you forward for physiography (indeed the latter is a contraction of the former). You can verify both assertions.[1][2] Vassyana (talk) 18:59, 28 November 2007 (UTC)

Well, by some definitions. :^)

I work with a bunch of geologists who would take issue with those definitions, as they don't include the geological sciences. I also personally know of numerous geologists who scoff at the idea that physiography is the 'new' term for geomorphology. They insist that physiography is a broader term for the geomorphology of landmasses which include other criteria as well, like flora and fauna, neither of which has anything at all to do with geology. Geography is more in line with spatial representaions of the Earth's surface, while geology is one component in what makes up the surface. Geography doesn't really care that much what makes up its contents. Also, much like our discussions over what the various disciplines call "primary sources", the same problem exists with "physiography". Different disciplines classify it differently, though there are similarities and even overlap. wbfergus ^Talk 20:30, 28 November 2007 (UTC)

Taking the side conversation to user talk. :) Vassyana (talk) 08:09, 29 November 2007 (UTC)

For anybody interested, I started on an article (right now it's still basically just a list) of the different physiographical regions. It still needs a lot of work, including references, but just going through trying to get all the appropriate wikilinks is time-consuming. I also might have to wait on the references (at least the ones for the maps) until January when I'm back in the office, unless I can get a friend of mine there to get into my office and add them for me. I think he's on vacation for a couple weeks though, so I'll play that by ear. I know it's hard telling if any of it is OR or not without citing any references, but I'd appreciate any feedback on what I have so far. wbfergus ^Talk 15:07, 2 December 2007 (UTC)

There are some exceptions that have been consistently applied in WP, related to primary sources. For example, the plot summaries in articles about films. See for example The_Joy_of_Sect#Plot, a Simpson's episode. When I asked in the FA review about possible violation of NOR, I was told by Wikiproject films members that plots can be easily verified by watching the film (the primary source) and thus are not in violation of NOR if well written and no attempts to provide interpretation are made. ≈ jossi ≈ (talk) 16:52, 28 November 2007 (UTC)

Yes. This is consistent with most previous interpretations of WP:NOR w.r.t. plot summaries in general, as also recently expressed here. ... Kenosis (talk) 17:47, 28 November 2007 (UTC)

I fail to see why jossi considers plot summaries to be an exception, unless he thinks summarizing is original research. If that's what he thinks, the solution is to go to the Wikipedia server farm and pull all the plugs. --Gerry Ashton (talk) 17:50, 28 November 2007 (UTC)

It appears that the WP users involved in that type of article have worked it out already and merely communicated the particular interpretation of policy to Jossi. I don't necessarily think Jossi, or anyone else, should be expected to know this unless he was actively involved in and familiar with that type of article or had asked one or more users who are already familiar with how it's been interpreted in that area of WP. ... Kenosis (talk) 18:07, 28 November 2007 (UTC)

I'm going to defend jossi's concern here, for one very good reason: about once a month, on average, I encounter some eager newbie who proclaims that primary sources should not be used in articles. As a result, there is a perpetual series of discussions explaining why a summary of a novel, play, television episode, etc. is not original research; you can't simply refer them to the policy, since the policy changes just enough that I can't be certain whether that is allowed, forbidden or not mentioned at any given moment. Further, if these newbies need help to understand why a plot summary is not original research, then sure as the sun will rise tomorrow they won't grasp the Tao behind ignore all rules -- which trumps that. (P.S. Now that I've argued that, I'll admit that there are also times where the best thing to do is to use secondary sources in plot summaries: for example experimental fiction like Finnegans Wake.) -- llywrch (talk) 22:08, 29 November 2007 (UTC)

Slim as I said higher up this page, I think it depends on whether one understands "and make no analytic, synthetic, interpretive, explanatory, or evaluative claims" to implicitly include "and make no analytic, synthetic, interpretive, explanatory, or evaluative claims [about the information found the primary source]". I take it to mean the latter. But if one does not and takes it to mean that no primary source can be used to make theses types of claims then out go many of the genocide definitions, because many of are sourced from their primary sources and all of them make such claims. --Philip Baird Shearer (talk) 18:13, 28 November 2007 (UTC)

Can you give an example of a genocide definition that is sourced to a primary source? Don't you mean sourced to an academic? How could an academic be a primary source on genocide (unless he somehow took part in a genocide he was writing about)?

Also, the addition makes no sense to me: "about the information found [in] the primary source." No, it's that primary sources should not be used to make analytic claims about anything. Primary sources should simply be used to show what the primary sources are saying. SlimVirgin ^{(talk)(contribs)} 03:45, 29 November 2007 (UTC)

Where the topic is Aristotle's Physics, Aristotle's work is the primary source; where the topic is For Whom the Bell Tolls, Hemingway's work itself is the primary source; where the topic is intelligent design, the works of leading ID proponents are the primary sources; where the topic is evolutionary theory, the original research of biologists, cladisticians, etc. are the primary sources (unless the point is evolutionary theory itself, in which case the primary source is Darwin or perhaps another original source of evolutionary synthesis); where the topic is the Book of Ezekiel, the scripture is the primary source. Where the topic, or thing being written about in a given part of an article, is the definition of genocide, the published definitions are the primary sources. Only if the topic is a particular act of genocide or set of acts of genocide, is an eyewitness account a primary source. ... Kenosis (talk) 04:57, 29 November 2007 (UTC)

Aristotle's work is the primary source for Aristotle's work, yes. But that doesn't mean that geographers are primary sources regarding the work of geographers. I'm not sure that's meaningful thing to say even. John Smith, the geographer, is a primary source for John Smith's work on geography, and Joe Blow's work on genocide is a primary source for his work on genocide. But it's stretching the definition to say that the published definitions of a word are primary sources. Yes, I see what you mean, but I think you're left with a meaningless definition that way. Do you have a source for what you're saying BTW? (We need a primary source for the definition of primary source. ;-)) SlimVirgin ^{(talk)(contribs)} 08:07, 29 November 2007 (UTC)

Suppose a reliable author observes some phenomena with his own eyes, and arranges for a reliable publisher to publish his analytic claims about them. Since the claims are not the result of reading and analyzing primary sources, the publication is a primary source, not a secondary source. A Wikipedia editor then incorporates the claims into an article, with no alteration, and with proper attribution. A primary source has then been (properly) used to make an analytic claim.

Perhaps SlimVirgin interprets "analytic...claims" to mean analysis of the sources by the Wikipedia editor. Perhaps SlimVirgin thinks the word "analytic" does not apply to analysis on the part of the author of the source. I think such an interpretation is far from obvious. --Gerry Ashton (talk) 04:07, 29 November 2007 (UTC)

If a newspaper reporter, for example, happens to observe an event, writes about the actual event and adds analysis to it, and his article gets published, that article would effectively be a secondary source. The reason is that the reporter with his analysis, along with possible fact checkers at the publisher, have taken the raw primary information ("2000 people attended the rally") and added their analysis to it ("it was the largest rally since 1976"). Wikipedians would then be able to use this analysis, as it is a legitimate secondary source. Crum375 (talk) 04:39, 29 November 2007 (UTC)

Crum375 tries to claim there is no such thing as a primary source that contains analysis. I disagree. --Gerry Ashton (talk) 05:05, 29 November 2007 (UTC)

It depends on the nature of the topic. If the topic involves analysis, the original source of the brand of analysis under discussion is the primary source of that brand of analysis. If it's a particular political theory or scientific theory, the original source of that theory is a primary source. If the discussion revolves around evidence that supports or contradicts a particular theory, the published research of particular studies or experiments are primary sources; if it's about a motion picture, the original published film is the primary source -- etc., etc. ... Kenosis (talk) 05:27, 29 November 2007 (UTC)

[Note:the following comments were posted earlier in response to the first query about "genocide definitions"]

No, the definitions at genocide definitions are all taken verbatim from their respective sources. No additional analysis or interpretation is added in that WP content. If this is being widely misinterpreted, then simply add the words "... about the information found the primary source" ... Kenosis (talk) 18:19, 28 November 2007 (UTC)

It should additionally be mentioned that the situation is utterly misrepresented as related. Only two of the sources are primary sources as defined in Wikipedia. Of those two, both are related in a secondary source as important and notable early definitions. Since the list depends on secondary sources, even for the inclusion of the two primary sources, how primary sources are treated would have zero effect on the article. Vassyana (talk) 18:27, 28 November 2007 (UTC)

I disagree, take for example "Genocide is a form of one-sided mass killing in which a state or other authority intends to destroy a group, as that group and membership in it are defined by the perpetrator." (The History and Sociology of Genocide: Analyses and Case Studies, Yale University Press [the primary source for this quotation]). This is a fundamentally different definition from that of the CPPCG and is frequently cited when people wish to use a broader definition of genocide than the legally defined one used by the courts which is based on the CPPCG. See the Genocides in history article or note the comment by the ICJ President Rosalyn Higginsin about the Bosnian Genocide Case "[because this judgement deals] exclusively with genocide in a limited legal sense and not in the broader sense sometimes given to this term". As I said this only becomes a problem if editors do not choose to read into the WP:PSTS clause "[about the information found the primary source]", something I think is implied by the sentence structure. --Philip Baird Shearer (talk) 20:48, 28 November 2007 (UTC)

I inserted the qualifying phrase here. ... Kenosis (talk) 21:19, 28 November 2007 (UTC)

Ahem,...

• If wbfergus' considers his example to be evidence of PSTS's confusing nature, then its not meaningful to suggest that his example is not admissible. Although I personally follow Crum375's line-of-view,...
  1. If someone says "policy" is confusing, then the first reaction of policy writers should be to accept that is in fact the case.
  2. And if he thought so, then others could think so as well.
  3. And don't the diverging opinions of whether it is/isn't primary research only demonstrate that he has a point?
• Its downright impractical to suppose that each and every case of primary blah blah needs to handled on a case-by-case basis. Basic rules, end of story. Only that with respect to PSTS, everyone clinging on to the section appears to have lost sight of what the base is (even though the conclusion of the section actually tells everyone what it is). Are the trees blocking the view of the forest?
• And... why is it anyway necessary for anyone to provide evidence of PSTS's confusing nature?
  The onus of providing evidence for the assertion that a differentiation of PSTS is necessary in NOR lies with those making that assertion. Not the other way around. Such is the way of the wiki.

-- Fullstop (talk) 20:40, 28 November 2007 (UTC)

In an effort to help clarify my position of my example above, I think the best way will be actually try to create some articles and add to others, getting it all tied together (without uploading the maps, which are probably copyrighted). Maybe after I get that done, either it will become more clear to me, or more clear to others, if and/or where problems may exist. Right now, I've just been tossing the problems around in my head, as the maps themselves are basically primary sources, being original work created by someone close to the subject. But, perhaps they could also be considered secondary sources, as they were also published, revised, and republished numerous times, and are now the basis for others 'work' on physiography. If when I get done it decided that there is a problem, well they could always be deleted. I'm not in the office anymore this year, so I'll send an email to a friend of mine in the office to help out as well, since he can get into my office where all my files are at. wbfergus ^Talk 11:18, 29 November 2007 (UTC)

In my opinion, PSTS requirements would be fully met by a statement in the body of the text on the lines of "physiographic regions based on maps by Edwin Raisz, A. Lobeck, and Guy Harold-Smith", and a secondary source showing the significance of these maps. That way, the primary source is used for facts that are obvious from the source, and the notability and credence of the source is shown by an independent secondary source. .. dave souza, talk 11:37, 29 November 2007 (UTC)

The onus of providing evidence for the assertion that a differentiation of PSTS is not necessary in NOR lies with those making that assertion. Not the other way around. Such is the way of the wiki. ≈ jossi ≈ (talk) 03:29, 29 November 2007 (UTC)

so much for perversion. "I'll make something up, you prove me wrong." Riiiight. -- Fullstop (talk) 14:51, 29 November 2007 (UTC)

No. It's more like when something has essentially been part of policy for over two years and eight months, the onus is on those trying to change a long standing status quo. Vassyana (talk) 21:49, 29 November 2007 (UTC)

I once got paid as a student to take part in some psychology research, where you had to sit in front of a computer that was flashing up words, and you had to hit yes if the word existed, and no if it was a made-up work. The speed of the words on the screen got faster and faster, and after 10 minutes, you started having no idea whether a word existed or not. They'd flash up words like "the," and you'd think "no, that's nonsense," then you'd think, "hang on, I think I've seen that word before." I'm starting to feel like that about "primary source." What the hell is a primary source?? ;-D SlimVirgin ^{(talk)(contribs)} 08:15, 29 November 2007 (UTC)

What would you like for it to be? Raymond Arritt (talk) 08:20, 29 November 2007 (UTC)

heh. -- Fullstop (talk) 14:51, 29 November 2007 (UTC)

"Three pounds of flax." -- llywrch (talk) 22:13, 29 November 2007 (UTC)

Where's the beef?

The PSTS distinction is commonly used as a useful reminder to take care to seek independent sources of verification for analysis, conclusions or notability of information from sources close to the subject being written about. As shown in the example posted here earlier. The terms are widely understood and used in common definitions, as the library links, but are not hard and precise. Some editors evidently find a conflict with more technical definitions, and find cognitive dissonance between their perception that rigid definitions rule out valid sources, and the looser definitions and negotiated consensus that commonly works in articles. The snow clearing at airports example shows a case where this policy has NOT been used to rule out a source, and where part of the sentence is unsupported by the citation. Are there any examples where this policy has actually caused a problem by being implemented to reject a good source, rather than a case where it's not been rigidly implemented? ... dave souza, talk 09:21, 29 November 2007 (UTC)

I don't have a case that fits Dave souza's criteria. If such a case existed, I imagine we would find that the source was not rejected entirely, but rather, certain passages were rejected on the basis that they were explaination, analysis, etc., within a primary source.

I suggest that we don't have to find such a source to warrant clarification of this policy. There may be a chilling effect, where editors would have used a good primary source, but didn't, because of this policy. There may also be cases where editors could have supported part of an article just fine with a primary source, but thought it wasn't allowed by this policy, so expended unnecessary effort to find a secondary source. --Gerry Ashton (talk) 16:18, 29 November 2007 (UTC)

It's meant to have a chilling effect on editors using primary sources for more than statements of fact, and encourage taking care to avoid undue weight as below. My suggestion at #How to use primary sources without original research was aimed at giving a way forward for such cases. .. dave souza, talk 18:32, 29 November 2007 (UTC)

A big problem with this policy and the discussion about it is vagueness about who exactly is writing statements. When Dave souza says "it's meant to have a chilling effect on editors using primary sources for more than statements of fact..." it isn't clear who is making more than a statement of fact, the cited author (which the Wikipedia repeats), or only the Wikipedia editor (which would indeed be original research). --Gerry Ashton (talk) 18:40, 29 November 2007 (UTC)

Although adding details to an intro paragraph is not good style, I believe the addition in the example should not have been entirely deleted. Instead, it could have been summarized into something like ", although this hypothesis has been subsequently questioned <ref>citation</ref>". The source seems perfectly appropriate to support a brief mention that an alternative hypothesis is under discussion, it's reasonable to mention this in an intro paragraph that describes the principle hypothesis. I simply don't see why the fact that it's a primary source makes it unreliable for this purpose. Best, --Shirahadasha (talk) 17:01, 29 November 2007 (UTC)

Point taken, but in the context of the feedback and the retention of the information in a section in the article, I agreed with the removal from the lead of the information I'd added... dave souza, talk 18:32, 29 November 2007 (UTC)

Use of primary sources to establish notability and significance claims

I think there is some confusion between (a) primary sources cannot be used to establish a topic's notability, and (b) primary sources cannot be used at all (i.e., they are not reliable). In general it strikes me as reasonable to have a rule of thumb that primary sources should not be used to establish notability, although they're perfectly good sources in other contexts. It seems to me that the issue of what sources are appropriate to establish a topic's notability is a subject that should be covered by WP:N rather than WP:NOR. The issue simply isn't one of original research. One is indeed citing a source for the notability claim. The issue is whether the source is appropriate for the specific sub-topic of notability. It's important to keep in mind here that WP:N was never made policy. It's only a guideline. Wikipedia doesn't strictly require notability at all, only verifiability and sourcing. I think the community should decide that notability should be required as policy before details about the sourcing of notability are made into policy. So long as notability itself is not policy, turning details about its sourcing into policy makes no sense. Best, --Shirahadasha (talk) 16:50, 29 November 2007 (UTC)

It's not just notability, it's the implied suggestion of significance in relation to the subject, effectively producing a synthesis that gives undue weight to what may be an extreme minority viewpoint. A primary source may well be useful for a statement of fact, but care needs to be taken to avoid introducing misleading impressions, even by inadvertent OR.. .. dave souza, talk 18:25, 29 November 2007 (UTC)

I agree, but the possibility of original synthesis has nothing to do with the degree primariness or secondariness of the sources you are synthesizing. I would argue that you have to be more careful about injecting OR with secondary sources, since they are more loosely tethered to the verifiable facts, and more easily manipulated in subtle ways, whereas manipulation of primary sources is obvious and easy to detect.

Shirahadasha makes a good point. Notability concerns are one of the main purported justifications for PSTS. Certainly, citation in secondary sources is a very good criterion for notability. But as WP:N says, once notability is established for a topic, you don't need to re-establish notability for every single sub-topic, claim, and bullet-point. Articles should be comprehensive about notable topics, and should use all reliable sources available: primary, secondary, tertiary, quaternary, or whatever. Once the notability of the general topic is established, the same standard applies for all sources: reliability. COGDEN 18:34, 29 November 2007 (UTC)

Given that notability concerns are guidelines only, and are some of our worst and most destructive guidelines at that (being used capriciously in A7 speedy deletions, the single way we WP:BITE the newbies most these days), I am loathe to edit a core content policy to be better compatible with a bad guideline. I'd rather make the core content policy sane (which, I continue to maintain, does not require PSTS to do it) than try to prop up bad policy with it. Phil Sandifer (talk) 18:44, 29 November 2007 (UTC)

The use of secondary sources as a benchmark of notability is explained in WP:FRINGE.

Which, until not long ago, that page succeeded in doing without any "help" from any editors of this talk page.

-- Fullstop (talk) 20:05, 29 November 2007 (UTC)

I agree with dave souza's point above that there is an issue of WP:SIGNIF as well as WP:N. I would also agree that it's reasonable to have a general rule of thumb that primary sources are typically not usable to establish that an opinion or point of view is significant. However, I would put the discussion of sourcing used to establish significance in WP:SIGNIF for the same reason I would put discussion of the sourcing needed to establish notability in WP:N. I don't believe either properly involves WP:NOR. Research and opinion doesn't have to be either notable or significant to be non-original. Currently, WP:SIGNIF is simply a redirect to WP:N, reflecting a current community consensus that the two should be treated the same way. I personally think this is not a good idea. The idea of significance is a critical concept in Wikipedia policy which really needs to be carefully defined and explained in its own right. I think content intended to get at this concept really belongs there, and WP:SIGNIF should be expanded to include it. Best, --Shirahadasha (talk) 05:58, 30 November 2007 (UTC)

Note: Here "significant" refers to the phrase"significant points of view" mentioned (but not defined or explained) in WP:NPOV. There seems to be a number of Wikipedia policy pages that attempt to get at this vital concept indirectly, yet none directly defines or explains it. WP:FRINGE is an attempt to describe what is not significant. We're discussing putting the sourcing for significance in WP:OR. However, we never actually explain what a "significant" POV is or how one determines whether a POV is significant or not. We don't discuss the role of the significance concept in Wikipedia policy or directly address its special sourcing consderations in the context of a discussion of the concept itself. I believe the concept of "significance" should be explained. I suggest WP:SIGNIF is the place for that explanation, including a discussion of the kind of sources needed to establish significance and special considerations about the sourcing of significance claims. The concepts of notability or significance, including sourcing issues related to those concepts, do not belong in WP:NOR. Best, --Shirahadasha 18:27, 30 November 2007 (UTC)

Case study: Philippe Casado

I wrote this article (from the article at fr and Babelfish), and added to the end of it a section about doping allegations. Obviously, for a professional cyclist, credible doping allegations are significant and belong in the article.

The problem is, the accusations weren't made directly - Greg LeMond spoke of "a former teammate" who left one team (Gan) for an Italian team that would provide doping products, and died of a heart attack shortly thereafter. Now, it is trivial research to look at LeMond's Gan teammates at the time and see that only one left for an Italian team and died shortly thereafter - Casado. This is completely verifiable, followable by anybody without specialist training.

Now, obviously the issues here aren't entirely primary sources - or even necessarily at all, since the article cites a secondary news account of LeMond's comments. But the sort of synthetic claim made here seems to me similar to the sorts of claims that (I think) we all agree can be safely made about the plots of movies and the like. And it gets at the problem at the heart of NOR - that there is absolutely nothing resembling a bright line distinction between OR and proper summary.

To my mind (obviously) the Casado article sits firmly and safely on the good side of the OR line - the claims are easily checked, the synthesis is obvious. But how do we go about tuning that distinction?

To my mind, our best bet is still, essentially, a social one - that is, claims that do not attract attention as OR are assumed not to be OR. But that would need to be paired with some sort of social norm that notes that you should not accuse something of being OR unless you have genuine skepticism that it is an obvious and indisputable conclusion.

What do other people think? How can we distinguish acceptable synthesis from unacceptable synthesis? Phil Sandifer (talk) 19:11, 29 November 2007 (UTC)

We cannot, and we should not try. We trust editors to use common sense and not do anything detrimental to WP, and this trust should extend to every sphere of editing process. That this trust is misplaced in some cases should not cause us to become generally distrusting. -- Fullstop (talk) 20:19, 29 November 2007 (UTC)

Agree. Recruit good editors and trust them. Let the editors involved decide whether the synthesis is reasonable or excessively original. --SmokeyJoe (talk) 20:47, 29 November 2007 (UTC)

I also agree, I think - local consensus and social control is going to be far more effective than a rigid policy. The problem is that we do have people who endorse rigid policy above all else - which is where we run into problems. What I'd like to see is a phrasing that encourages social norming here - something that encourages to challenge things they think are non-obvious or disputable conclusions, but also discourages them from challenging things for the sake of the rules alone. And that's, I think, where a lot of the problems with this page could quickly be solved - if we effectively discouraged treating NOR as an end in itself. Phil Sandifer (talk) 21:13, 29 November 2007 (UTC)

Well, until we can monitor the front door to the project, I suspect that we will always have a problem with people who want to rigidly enforce policy. I've been wrestling with a similar problem: providing a summary of the events of an important battle in Ethiopian history. The primary source I would use narrates the events quite clearly until the middle of the battle the author starts a new paragraph about how one side has won the battle; when I read that passage, I felt that the words "a scene missing here" needed to be added in the text -- yes, just like in the Monty Python sketch. Now, I can reference to traditional Ethiopian warfare tactics, & fill in the gap with a "this is likely what happened here" explanation, which would not be novel or unusual -- but I'm not comfortable with that, & until I can solve this problem the article is likely to remain a stub. What would help me in this case -- & I believe would help in the case Phil Sandifer describes above -- is if we had a way to get an outside opinion from a veteran editor who could help explain that this is not OR. I don't think opening an RfC would work because of the way this method is currently structured (e.g., it needs 2 editors to open), & because some eager-beaver newbie would see the RfC & slap {{Original research}} on the article -- which would defeat the purpose but succeed in annoying one & all. -- llywrch (talk) 22:37, 29 November 2007 (UTC)

I agree - an improvement in the culture we have surrounding the act of editing (as opposed to just writing) would only be a good thing. Phil Sandifer (talk) 23:57, 29 November 2007 (UTC)

Ah yes. The commonality then is "everything in moderation," here avoiding over-precision on policy pages ("spirit, not letter"). -- Fullstop (talk) 23:55, 29 November 2007 (UTC)

I think this has more to due with reliability than with OR. In terms of OR, I'm not a cycling fanatic and don't really follow the sport, but it's obvious to me who is being referred to with minimal checking. The blatantly obvious to non-experts has long been an accepted exception to the rules, and is even explicitly included in the language of V and NOR. However, it's questionable whether an implication is acceptable from a very visible and acrimonious crusader against doping who all but accuses the entire sport of doping making an uncorroborated accusation on a hot topic that is widely covered by the media and other sources. There would be less question if the accusation was in the accuser's article discussing his anti-doping crusade. Vassyana (talk) 22:30, 29 November 2007 (UTC)

Second case study: Jacques Derrida

A second case study, to talk about some related issues: Jacques Derrida is, without a doubt, one of the most important philosophers of the 20th century. His work, however, is deeply impenetrable - exactly the sort of primary source we don't want non-experts playing with. Unfortunately, secondary sources are pretty darn dense too, because they're all academic publications aimed at academic audiences, and few if any make a real move to be clear to non-specialist readers. Some tertiary sources exist, but, frankly, they're not very good, and do not capture what Derrida is doing well.

The way people learn about Derrida is through coursework - coursework where Derrida's works are assigned, and, through class discussion and the leadership of the professor, some understanding is come to. In other words, understanding of Derrida is not something that is, in the academy, communicated particularly through written works, but through an oral tradition. This is not surprising - Derrida is a graduate-level topic, for the most part, and graduate courses in the humanities are very primary-source focused, as they should be.

My question is not how to go about writing a Derrida article - we've done a decent job of that. We did it because people who know about Derrida have written from their knowledge, and provided enough sources for readers to familiarize themselves with the literature should they choose. But there are sections of the article - good sections - that could be ripped apart if somebody decided to apply WP:NOR and WP:CS rigidly to the article.

That's the problem I have. Especially since Derrida is far from alone on this - most topics studied by graduate students in the humanities face the exact same problem - high notability, few to no accessible secondary or quality tertiary sources.

This policy needs to be formulated in such a way that productive work on these topics can continue as it has. That is to say, this policy needs to reflect actual good practice in the article space. What can we do on this front?Phil Sandifer (talk) 19:11, 29 November 2007 (UTC)

I worked on that article a couple years ago, and actually, I'd say that in this field, everything is a secondary source, as well as a dense, highly-technical primary source. Almost everything Derrida wrote, for example, is a secondary criticism of some prior philosopher or struc[turalist, as well as a seminal primary source on deconstruction. You can't separate the primary and secondary aspects of his work, because they are one and the same. Plus, you can't really write about Derrida in either a strictly primary or secondary way. In fact, when writing about Derrida, using a primary-secondary distinction is almost laughable, kind of like classifying sources on nihilism as good or evil. The whole primary-secondary framework goes out the window, leaving us with the idea of original research, which is where the focus should be anyway.

To write about Derrida without doing original research just means that you don't make any statements about Derrida's works, or works about Derrida, that would not be considered obvious to a highly-informed reader of Derrida. This is one example, too where the "reasonable, educated person without specialist knowledge" requirement has to go out the window. No such person would be able to verify the accuracy of a citation to Derrida, any more than they could verify a citation to a nuclear physics article, unless it is a direct quote. COGDEN 20:13, 29 November 2007 (UTC)

These are all good points. SlimVirgin ^{(talk)(contribs)} 20:54, 29 November 2007 (UTC)

An article on Derrida (or on any author for that matter) cannot use that author's own statements as a source to establish anything about the author unless those statements are explicitly (auto)biographical. As always, source type doesn't matter (not that a distinction, as Cogden points out, applies to Derrida). What matters (with respect to NOR) is that the statements be accurately reflected.

Incidentally, there is one sentence that I (non-Derrida person) immediately picked out in the Derrida article as blatant OR:

"It can plausibly be argued that with this commentary Derrida had already posed the basis of his whole path of thinking.^[11][12]"

The "plausibly" is evaluating the two cited sources. The phrase "can plausibly be" should be replaced with "has been."

-- Fullstop (talk) 21:15, 29 November 2007 (UTC)

That's probably just poor phrasing, though, and doesn't get at the issue I'm talking about. I say probably just poor phrasing because one of the sources seems to be an interview with Derrida, and the subject is such that the phrasin ghtere could actually be accurate - that is, that the argument advanced may be that the interpretation is plausible, not that the interpretation is true. I don't have easy access to the sources, though, and odds are better that you're right - I bring this up only to reiterate my point - one made very succinctly by Cogden as well - for writing about Derrida, the "without specialized knowledge" criterion is not a usable one. Phil Sandifer (talk) 21:20, 29 November 2007 (UTC)

Derrida is the same as other dense and difficult but highly notable philosophers. I've yet to encounter a notable difficult-to-understand philosopher whose work has not been explained in a relatively simplistic manner by textbooks and other course materials. Hegel and Husserl are two prominent examples. The phenomenology of Husserl is an incredibly complex topic, usually studied by graduate students, but undergraduate level "summaries" of the difficult topic are far from scarce. Derrida is no different. Vassyana (talk) 21:54, 29 November 2007 (UTC)

Sure. It's just that those summaries are, at least from academic vantages, not very good. I mean, I certainly don't want a Derrida article written primarily from Derrida For Beginners. Which is not to say that the book doesn't have its uses - but no responsible academic would say it provides sufficient basis to use as a major source in writing a summary of Derrida. Phil Sandifer (talk) 22:09, 29 November 2007 (UTC)

The reason they're "not very good" is because they are (relatively) exceedingly short summaries of highly complex topics. However, they're very good for providing a rough overview and introduction to the topic. Wikipedia is written for a general audience and requires a pragmatic article length. A very well written Wikipedia article should be roughly equivalent to those "not very good" sources. Wikipedia simply isn't the place to get into complex lengthy graduate level explanations of topics. Those "poor" sources are written by academics for post-secondary academic instruction and subject to a very competitive market. I'm sure they (as a whole) know better than you or I what is the appropriate way to summarize the topic. Of course academics who know the topic inside and out are going to criticize the "oversimplification" of the topic. That's a fairly universal reaction across fields regarding summary and introductory sources. Vassyana (talk) 22:21, 29 November 2007 (UTC)

Yes - Wikipedia is a tertiary source. But if we write from already poor sources we're going to get even worse sources. And, I mean, I'd like to be, as Jimbo's slogan goes, better than Britannica. After all, part of why Derrida for Beginners and other sources are so mediocre is that they try to cram a lot of information into a short space. Wikipedia is not paper - we can afford to actually give a good, thorough, not simplifying and inaccurate overviews of topics. That's one of the reasons we're better. Phil Sandifer (talk) 22:33, 29 November 2007 (UTC)

Wikipedia isn't paper, to be sure, but it's also not intended for articles to be ridiculously long. Even a very long article (by Wikipedia standards) would generally be shorter than those summary sources. Again, I don't think those sources are "poor", but rather necessary simplifications of the topics. We're not going to be able to do much better without making an article incomprehensible to the reader or drawing on sources that can't be verified by anyone without extensive expert knowledge. Both such circumstances are highly undesirable. Providing the (over)simplified summary should be the practical goal, along with providing external links and further reading recommendations for those seeking a more in-depth understanding of the topic. Vassyana (talk) 23:19, 29 November 2007 (UTC)

Since the number of our articles is, for practical purposes, unlimited, there is no reason we cannot have both over-simplified introductory articles and more advanced articles appropriate for readers with more background. — Carl (CBM · talk) 23:27, 29 November 2007 (UTC)

Point taken. I'd imagine in your conception that this would basically work like the normal subarticle/spin-off process? That is, the more advanced articles addressing specific subpoints in depth? Would you agree our primary goal should be first to create decent overview articles on the main topics? How would you address the problems of verification and identifying original research inherent in such complex topics? Vassyana (talk) 23:44, 29 November 2007 (UTC)

One thing that would be helpful is to ditch the "non-specialized audience" requirement - it's not helpful on the easier articles either, honestly. What would help there is a push towards phrasing things such that a non-specialist audience can understand them - but that's a problem across a lot of topics that does not seem to me to be inherently related to OR. I think we can write a better basic article on Derrida from sources other than the tertiary sources too. Phil Sandifer (talk) 00:05, 30 November 2007 (UTC)

Vassyana, I would be glad to discuss that question, but I don't think this is the right place, since that issue runs much deeper than NOR. I think about it quite often as an editor of technical articles. If you would like to talk about it, I'll volunteer my talk page for everyone to use. To give a very quick answer, I do think a spinoff process is reasonable, although sometimes it is faster to dig a tunnel from both ends towards the middle.

Phil, I think the non-specialized reader requirement is important as an anti-crank tool, which I can explain upon request. — Carl (CBM · talk) 00:23, 30 November 2007 (UTC)

Cranks tend to have a number of other problems, including m:MPOV and complete fucking insanity. I suspect we can soften this policy to better govern normal cases without losing our ability to stop the utterly insane. Phil Sandifer (talk) 02:45, 30 November 2007 (UTC)

A serious problem with using more basic summary works is that they represent one highly-disputed POV. Derrida himself, and many other commentators, opposed attempts to simplify his writings, and the POV is that these simplifications are simply wrong. And reading some of these "Deconstruction for Dummies"-type books, I tend to agree. Derrida's work, I think, is just not simplifiable consistent with WP:NPOV, which is the most fundamental and inviolable Wikipedia policy.

I don't think there is anything wrong with articles written for more advanced readers. There are hundreds of them. The vast majority of mathematics and physics articles are way behond the non-specialist reader, and sources in it could never be verified by someone without very specialized knowledge. For example, consider the article Quantum chromodynamics. It's a significant and notable topic, but nothing in this article is comprehensible by anyone other than a trained physicist. Moreover, you could not expect anyone other than a physicist to be able to verify any of the references (primary, secondary, or otherwise). Furthermore, there are no "Quantum Chromodynamics for Dummies" books in existence, and even if such a book existed, it would be directed to physicists who already have specialized training in quantum field theory. The "non-specialized reader" requirement just isn't feasible in all cases. So either we drop that requirement, or we effectively ban articles on a large number of notable topics. COGDEN 18:34, 30 November 2007 (UTC)

The section on Derrida's life is essentially uncited. But if verfiability is not an issue and people are not questioning the content, I don't see what the problem is. If there are statements in the article that there is any doubt about, request citations and go from there. I personally might want to see a bit more secondary or tertiary citation for some of the insights into his work that draw directly on quotes mined from Derrida's own work (primary sources). But on the whole I don't see a problem, quite frankly. The overwhelming majority of the material in the article appears likely to be verifiable from sources other than Derrida's own statements. If anything in there isn't verifiable, then follow WP:V. WP:V#Burden_of_evidence says essentially "if it's not verifiable, take it out." So if looking for a policy to soften up, maybe WP:V is another good candidate at the moment.

I notice there are some problems with the "criticism" section at the moment, which is very common in philosophy-related WP articles. To the extent that most secondary sources are also obscure w.r.t. Derrida's work, tertiary sources would appear to be helpful in bringing this difficult-to-write article into a better written and yet-more-useful article in the future. The online Stanford Encyclopedia of Philosophy has an article on Derrida, as does the Routledge Encyclopedia of Philosophy (both the full version and the brief version) and also the Macmillan Encyclopedia of Philosophy (the recently published second edition, not the 1967 one). Internet Encyclopedia of Philosophy also has one that could be helpful as "grist for the mill".

Incidentally, if anyone wants to see some real OR related to Derrida, do check out the current version of deconstruction, starting right in the first paragraph. ... Kenosis (talk) 04:59, 30 November 2007 (UTC)

The solution is not to turn to further-removed sources that are easier to read; it's to find someone who is already familiar with Derrida and the most relevant secondary sources, and get that person to help with the article. Although Wikipedia is the encyclopedia anyone can write, that does not mean that anyone can write an article about Derrida. The same could be said for numerous topics, from philosophy to engineering to music to popular culture. Our goal is to write an encyclopedia, not a compendium of grade-school book reports. — Carl (CBM · talk) 13:19, 30 November 2007 (UTC)

Yes. Exactly. And while the deconstruction article could use its sources more clearly, and though it's more SPOV (sympathetic point of view, not scientific) than NPOV for some sections, it's also more faithful to deconstruction than most summaries of its length that I've ever seen, and frankly doesn't bomb accessibility too horribly either. That is, though it has problems, and though the "deconstructionism cannot be defined" part should probably be moved behind a (however poor) attempt to define it, it is, frankly, one of the better tertiary sources I've ever seen on the subject. I would be loathe to lose its accuracy in favor of a rigid enforcement of NOR, especially as little to nothing in the article strikes me as particularly original. Which gets to my original point - given that deconstruction is, in practice, a pretty good article, how can we adjust this policy to reflect actual articlespace practice? Phil Sandifer 13:50, 30 November 2007 (UTC)

Also agreed with Carl on that. Only if there's a disagreement or doubt about the content does even WP:V "kick in", so to speak, as to providing inline citations. If there's no disagreement, for example, about the contextual material provided at the beginning of the section of Derrida's work, don't make it into an issue, which would IMO be nothing more than wikilawyering unless a specific issue is articulated on the talk page. If there is doubt about the contextual material provided, start by consulting the available tertiary sources, several of which are of the highest level of quality, and work from there. Please do not use "Derrida for Dummies". or the Cliff's Notes for On grammatology ;-)

Another approach, as has been done with many topics that are highly technical or obscure, is to break it up into an introductory article and an advanced article. Some examples are listed at User_talk:Kenosis/Research#Basic_v._Advanced_articles. Obviously there's already some content disagreement in the "criticisms" section. Several of the available tertiary sources on Derrida are among the best in the business, and at some point in time they ought be consulted in any serious attempt to resolve oustanding issues-- assuming there's adequate interest in resolving them. But the far more noticeable problem, IMO, is not with the article on Derrida, but with the article on deconstruction which appears to be where the secondary and tertiary sources will be more useful if not absolutely necessary if the article is to be much further improved. ... Kenosis 14:22, 30 November 2007 (UTC)

To be frank, though, there is no need to adjust the policy in response to the fact that some articles are arguably quite good when citing directly to the primary source(s), as is the case in the Derrida article's section on his work. If there's an issue with the analysis, secondary and tertiary sources will be needed to support the conceptual assertions in the section on his work. IMO, it's pretty much that simple. ... Kenosis 14:39, 30 November 2007 (UTC)

Again, though - it would be very helpful if we could put some anti-rules literalists language into the policy to prevent crusaders and people who are seeking a ton of maintenance edits in pursuit of an admin run from tagging stupidly. this is where I'm unconvinced the "non-specialist" clause helps us - it is seemingly intended to help ward off cranks, and I think we can do without it for crank warding, and that not having it would make the policy better reflect reality. Phil Sandifer 15:32, 30 November 2007 (UTC)

Agreed about the need for counteracting excessive literalism in policy interpretation. I have seen, for example, numerous "drive-by" unreferenced templates and citation-needed templates that arguably are quite gratuitous. It's extremely easy to say "hey, where are the references as required by policy". But somebody has to do the actual work, after all. By placing the template, a user actually sets the WP:V and WP:NOR policy analysis into motion. When this happens gratuitously, it might be best to wait for an explicit statement of what the question is about the material, and lacking an explicit statement of what the concern is (beyond the assertion that a particular passage or section is not explicitly cited), remove the template rather than the material itself, But this is also dependent on the nature of the material. Jimmy Wales' statement about WP:V, quoted on that policy page, gives high priority to eliminating unsourced negative statements in BLP's. Although Derrida is no longer alive, an explicitly derogatory statement about him would immediately merit close scrutiny. This, as with many or most situations, necessarily involves factoring in more than one aspect of editorial policy in order to arrive at a reasonably sensible resolution. ... Kenosis 17:59, 30 November 2007 (UTC)

I think that all, or almost all, secondary sources on Derrida introduce POVs distinct from that of Derrida's work, and take on a life of their own outside the scope of the Derrida article. This is one of the hazards of critical studies, that every discussion of Derrida is a criticism of Derrida of one sort or another. Not necessarily opposition, but at least a different take on it. Many of the "Derrida for Dummies"-type books are actually ironic pastiches of Derrida. Using such secondary sources alone is insufficient, and you'd have to use them with great care, and support them with references to the primary sources—the exact opposite of what PSTS currently says. COGDEN 19:11, 30 November 2007 (UTC)

Drawing on Derrida to provide yet another interpretation/take on his writings and philosophy would be original research. You aptly demonstrate here exactly why there is a PSTS section in the first place. Vassyana 19:45, 30 November 2007 (UTC)

As to anti-literal-enforcement language. I don't think that's the answer. If the rule is broken, let's fix it, not authorize people to break it. We can fix the rule by doing two things:

1. Replace the rule that primary source citations must be "easily verifiable by any reasonable, educated person without specialist knowledge" with a requirement that any citation (primary or secondary) must be "verifiable by a person with an ordinary level of education and training in the field of the subject matter being discussed".
2. Treat primary and secondary sources under the same standard, since in technical and philosophical fields, the vast majority of secondary sources are more technical and specialized than primary sources, since they build upon the primary sources and assume them as a prerequisite. (Tertiary sources like textbooks might be less technical than the journal articles they cite, but why confuse things with a tertiary category anyway?)

COGDEN 19:11, 30 November 2007 (UTC)

1. Totally unenforceable suggestion, given the current situation. It would require a drastic change in Wikipedia, including a fundamental shift in the roots of the project, to allow for expert authority. Wikipedia in many cases does encourage experts to contribute, but still requires them to follow the same policies and practices.
2. I know this to be quite less than true, as a general rule, for philosophy and history. While some peer-reviewed papers can be horridly complex, those are almost exclusively limited to articles addressing very specific and technical points. (Even worse, the understanding and interpretation of primary philosophical and historical texts is almost unavoidably original research without secondary sources.) I also have a hard time imagining that secondary sources are more complex in most technical fields with which I am familiar. For example, books about particular programs and programming languages are usually far more understandable to the average person that the code and documentation (secondary and primary sources respectively).

I appreciate your thoughts as good faith ideas, but they are very counter to the grain of Wikipedia's fundamentals. Vassyana 19:45, 30 November 2007 (UTC)

I disagree with your interpretation of #1 - this does not give experts special privilege - rather, it changes our editorial standards away from judging articles from the perspective of people who do not know anything about them, and demands that our articles read as good to experts. That doesn't necessarily equate to credentialism. To my mind, Wikipedia is based on the assumption that the people who write articles know a lot about the subject, and that we don't have to check their credentials because they're going to cite sources, be NPOV, et. Phil Sandifer 23:14, 30 November 2007 (UTC)

It most certainly does give expert editors special privilege. Allowing for information that can only be verified by experts is certainly a huge change and most certainly according special privilege. If it's just a matter of making sure articles read well to experts as well as the common editor, they are existing mechanisms for asking for expert attention on an article. Similarly, if we're still relying on normal checks for NPOV, reviewing sources, etc, then the proposal is just fuel for trouble. It is either unenforceable or enforced credentialism. Without credential verification, it becomes nothing more than an excuse to hear more claims of "But I'm an expert!" and would encourage those with an agenda to cite obscure and difficult to comprehend sources relying on the proposed expert provision to push their POV. These are already concerns present in the wiki, though the former is much mitigated since a certain false credentials debacle. Vassyana 23:44, 30 November 2007 (UTC)

But the alternative contention - that non-experts can helpfully check information on complex topics - is ludicrous. And, given that topics are likely to attract expert editors, it seems to me that raising (or lowering, depending on your perspective) the bar is not a large risk. Sure, a crank might begin citing obscure and difficult papers - but said crank is unlikely to be the only expert on the topic editing the article, and if the crank is, odds are that the article is obscure enough that one bad editor will be able to effectively derail it with or without the protection of NOR. Phil Sandifer 00:18, 1 December 2007 (UTC)

Please don't misunderstand me as saying it's a change that shouldn't be made. I've taken no position on whether we should or shouldn't move in that direction in one form or another. I am simply pointing out that it is a very fundamental change in the wiki, and such a large change needs to be considered as such. Without some form of verification, the distinction just becomes a claim editors use to win arguments, justly or unjustly. A crank could control an article with claims of professional expertise drawing on esoteric and recondite references, because most "experts" on Wikipedia are dedicated amateurs, not trained professionals. (Upon thought, I think it would be within Wikipedia practice to allow for verification by those familiar with the field (without specifying professional training or expert knowledge). This would be akin to the allowance for foreign language works, though like that provision I'd imagine it'd be somewhat controversial.) Vassyana 22:45, 2 December 2007 (UTC)

Vassyana, when you talk about computer books about particular code, you seem to be talking about what I would call tertiary sources, such as textbooks, where this is true. For example, while the Haskell programming language standard (primary source) might be referenced accessibly (relatively speaking; not to the layperson) in the Programming in Haskell textbook referencing it (a tertiary source), a work in the Journal of Functional Programming (a secondary source) that references the standard will be far more technical than the standard. Then a second secondary source that cites the first journal article as a primary source will be more technical and specialized still.

Likewise in other fields, you can expect that a work by Derrida (as a secondary source) will be far more technical and specialized than the works Derrida references such as phenomenology works by Husserl (primary source). Similarly, Husserl as a secondary source will be far more technical and specialized than a work of Hegel (primary source) that Husserl cites.

I disagree that this will change anything about Wikipedia. This is exactly how Wikipedia currently works. Just look at all the math and physics articles, and see if you can understand them. There are thousands of them, and they are very valuable resources. Do we delete them? Plus, this practice is fully consistent with WP:V as it has existed for years, which requires verifiability to reliable sources, but does not say anything about the qualifications of the verifier. That the average layperson cannot verify the citations in a physics article is no problem, and nobody I know of has ever seriously complained about this in the history of Wikipedia. The only problem I forsee is when the intended audience cannot verify the citations, which is why I'm suggesting #1, which represents a strengthening of WP:V. COGDEN 23:57, 30 November 2007 (UTC)

Philosophers often draw upon other philosophers to support their claims or provide contrast. In nearly all cases, you could hardly call them secondary sources for it, with the exception of papers that specifically analyze and/or critique the work of others. To draw a parallel, an astrophysicist may draw upon special relativity for some of her claims, but that doesn't make the article a secondary source on relativity. Also, you're not very familiar with Hegel or Husserl if you think directly citing them is a good idea or easier to understand than secondary sources analyzing them. Like Derrida, they are notoriously easy to misunderstand and the in-depth study of those philosophers is the subject of graduate coursework. However, also like Derrida, the foundational concepts and contributions to philosophy can be (and are) related in a general summary fashion to undergraduate students.

An additional problem with direct citation of a philosopher's work is the propensity for original research (directly tying into PSTS I might note). It is nearly impossible to cite a philosopher's work without OR, except explicitly saying X said "Y" (which can still carry concerns of reliability and NPOV). This is because any such citation is another interpretation of their work. While cases like Derrida, Hegel and Husserl are obvious in this class, even more commonly understood philosophers still remain. Philosophers still argue the nuances, and in some cases even the essential meaning, of Aristotle and Descartes for example. Those cases which are plain to everyone obviously aren't problematic, because the plainly obvious to most/all (or "non-controversial") is a long-standing consensus-based exception to most content rules.

I'm not a science and math wizard, but there are some fairly intense and complex math/science topics on Wikipedia that are understandable (to me) as presented, despite my lack of in-depth subject knowledge. Can I understand complex equations and the like? No. Could I still easily verify such an equation by sight alone in a cited math text? Yes. Complex topics do not have to use language incomprehensible to anyone but experts, even if some information is difficult to understand for non-experts (such as complex equations). Additionally, your proposal specifically requires training in the field (credentialism), which is distinct from current practice. I'm sure many (if not most) of our in-house "experts" are interested amateurs, as opposed to trained professionals. Vassyana 22:45, 2 December 2007 (UTC)

This is not a proposal. This is the way it's done, and has been done for 7 years at Wikipedia. If you want to lobby to change this, then fine, but I doubt you'd get much traction. The purpose of policy articles "is to record clearly what has evolved as communal consensus in actual practice" (see WP:POLICY). Actual practice, as evidenced by thousands of math and physics articles, is not to have a "verifiable by any reasonable, educated person without specialist knowledge" requirement".

If you agree that philosophers are notoriously easy to misunderstand, then why would Derrida be less easy to misunderstand than Husserl when Derrida quotes Husserl?

Another question, if you don't think that philosophers discussing other philosophers are secondary sources, then what is a secondary source? Since it's really only philosophers who publish works discussing Derrida, that means that (1) you can't cite Derrida, and (2) you can't cite anybody who cited Derrida. Who is left? Also, do you really think that Derrida discussing Husserl is easier to understand than Husserl himself? It's not true, in part because when you read Derrida, not only to you have to understand Husserl, but you have to understand Derrida's take on Husserl. COGDEN 08:39, 3 December 2007 (UTC)

I disagree that it is common Wikipedia practice. At best, it is something permitted in a very limited area of the wiki (significantly complex math and science). Taking common practice as a whole, it is an exception not a rule. In philosophy, papers and books undertaking to specifically critique, analyze and/or explain a philosopher or philosopher are what would fall under secondary sources. Philosophical treatises citing other philosophers is just another primary paper. Something like In the Spirit of Hegel by Robert Solomon (ISBN 0195036506) would be a good secondary source on the phenomenology of spirit, for example. We should be relying on what others have published about the meaning and so on of a philosopher's work. We shouldn't be interpreting philosophical treatises directly. If the interpretation has been published before, we should be attributing the source. If it hasn't, or a reference cannot be provided, then it's original research. Vassyana 09:42, 3 December 2007 (UTC)

As they say, "it is the exception that proves the rule". And indeed, the rule is proven weak and invalid. Are math, science, and philosophy articles now the bastard children of Wikipedia? Do they live in the lawless frontier where where the rules aren't applicable? If so, we need to explicitly state this in PSTS. If such a statement in PSTS would fix the rule (from your perspective), it would be negligence not to do so. COGDEN 20:52, 3 December 2007 (UTC)

(outdent) Actually, the exceptions simply show that the rule is broad and not applicable in all circumstances. This characteristic of Wikipedia policy is well-considered and acknowledged. We even have policies that account for it (like say, Wikipedia:Ignore all rules). Since the rules are not meant to be all-encompassing and exacting, limited exceptions do not harm the strength or validity of the rule. Your implication that math, science and philosophy are exceptions as a whole is misleading (at best). It is a small minority of those topics that stand out as exceptions. We don't need to point out that exceptions exist, though I'm keen on including such caveats in areas of policy that are abused in a legalistic fashion. (For example 3RR includes a caveat that it's an electric fence, not a permissive limit.) Vassyana 22:54, 3 December 2007 (UTC)

We can't rely on WP:IAR to fix all the errors on policy pages. We won't have to ignore all rules if the rules are correct in the first place. Do you really think that it's better to have an flawed, but easily-fixable rule, fixed by IAR, than just to fix the rule? We have an opportunity to make the policy article conform to actual Wikipedia practice (see WP:POLICY). Why would we not take that opportunity? Is it because this page is an exception to WP:POLICY and WP:CONS? COGDEN 01:27, 4 December 2007 (UTC)

I'd agree with your initial statement. I'm sure you're aware that I agree there are things that can be done to help make policy more accurate and inclusive. However, I'd disagree about the assertion that IAR wouldn't be necessary for "correct" rules. I believe there's no ignoring the basic principles of policy, but that the policy itself will never be all-encompassing or otherwise "perfect". The community is simply too large and the content too varied to make "correct" rules that deprecate the need for IAR without delving into extremely lengthy and complex codification. I don't think policy should be built on rare exceptions and tiny minorities. Certainly, as I mention above, I wouldn't have much issue with something that accounts for the need for verification by those familiar with the topic in limited circumstances. That's just recording common practice. Allowing for a verification standard that permits incredibly complex material only verifiable by professionals is certainly a large step away from that and from established Wikipedia culture. Vassyana 01:59, 4 December 2007 (UTC)

As a general comment, I believe literalism/legalism (in relation to policy interpretation) is counterproductive. After all, the rules are principles, not a civil code. Vassyana 19:45, 30 November 2007 (UTC)

As a general comment, literalism/legalism in NOR is what you appear to be insisting on. -- Fullstop 20:23, 30 November 2007 (UTC)

Utterly false. My opposition to extreme simplification does not mean I support a legalistic or absolute interpretation of policy. What I mainly desire is a compromise between the various views espoused regarding PSTS. Vassyana 20:52, 30 November 2007 (UTC)

Vassyana, there is not only one shade of gray (yours) between black and white.

And, that shade of gray is, contra your insistence without accompanying reason, not good for the 'pedia.

And, it would also be enormously helpful if occasionally did not say "no." The frequency with which you do does not support your supposition that you desire "compromise." People who want compromise don't talk like you do; there is no such thing as "utterly false," just as there is no such thing as "utter truth."

-- Fullstop 23:31, 1 December 2007 (UTC)

My honorable Wikipedian, needs to read the archives and the long discussions on this subject to understand Vassyana's somewhat forceful assertion. ≈ jossi ≈ (talk) 00:52, 2 December 2007 (UTC)

"My" shade of grey isn't something I purely invented ex nihilo. It was an accommodation of the principal views espoused over several months of discussion. It was further revised to address concerns raised, as practical. It was a sufficiently accurate attempt in that regard, as it generated support from those engaged over the long-term on both "sides" of the long-running dispute (for example, both Jossi and COGDEN consented to the compromise). I have failed to say "no" on multiple occasions. For example, I have often revised drafts in response to concerns and criticisms that could be handled in the framework of compromise. I do admit I have said "no" repeatedly to untenable demands, such as your own extreme solution to the issue. However, that's hardly anything indicating a lack of desire for compromise. On the contrary, rejecting all-or-nothing solutions (while accommodating the root concerns as much as possible) is part and parcel of seeking compromise. Regarding your closing comments, it's perfectly common and acceptable language to deny false accusations with statements such as "utterly false". Such a forceful reaction is perfectly within the norm and doesn't speak to my interpretation of policy or my actions in seeking compromise. Vassyana 22:45, 2 December 2007 (UTC)

I have not only read the archives and long discussions on the subject, I have also participated in them.

And, as have many others, I have also repeatedly questioned the necessity of PSTs distinction, an answer for which has always been 'judiciously' evaded.

And, as have many others, I have also repeatedly questioned why it can't be dealt with on its own page (IIRC, one of those posts [not from me] was in fact addressed directly at Jossi, but an answer, as for all the other questions, never came).

So honorable Wikipedians, perhaps its time you took up the challenge and answered the questions. To the point.

-- Fullstop 01:27, 2 December 2007 (UTC)

To be very blunt, if you've read the archives as well as you claim, your question has been repeatedly answered. However, I will give you an explanation, so you cannot claim you've never found an answer.

The necessity of the distinction arose due to the abuse of primary sources for original research, including unintentional errors of that sort. The abuse was noted when the distinction originally entered the policy in early 2005. The abuse was still perceived as a concern when the policy language was further refined in mid to late 2006. Imposing a drastic replacement on 2-2/3 years of standing policy requires an impressive burden of proof and consensus. (Though, my desire for compromise is rooted in the currently controversial nature of the section.)

Many primary sources are near impossible to use without violating NOR, unless following the use of such a reference by a secondary source. For example, standard citation of Caesar's Gallic War (outside of following the use of a cited scholar) is severely problematic and OR on its face. It's among the better ancient sources, as scholars acknowledge it as a pinnacle of ancient reporting. However, it's also known to be a masterpiece of propaganda and to contain numerous pieces of inaccurate hearsay. Without published expert guidance, it's entirely OR to judge whether a passage is reliable (which is what standard citation essentially does). However, we are allowed under policy to simply say The Gallic Wars reads "X". (There may also be concerns of reliability and reflecting the body of reliably published work, but those are distinct concerns of WP:V and WP:NPOV).

A common counterargument is that non-reliance on primary sources is directly contrary to standard scholarship. However, that is more an argument for PSTS distinctions, than against. Standard scholarship's aim is to produce original research. Reliance on primary sources is encouraged to help ensure less reliance on the claims of others and to exercise the scholarly tools of analysis on the "root" material, both of which are contrary to the central principles of Wikipedia. Wikipedia is intended to summarize what others have said, not as a forum for scholarly analysis.

There is simply a broad and strong opposition to moving PSTS from the framework of this policy. As noted above, the distinction directly relates to this policy. Vassyana 22:45, 2 December 2007 (UTC)

Thank you for at least responding, even if does not answer either question. This failure may be attributed to your not listening.

As I said before, I have read the archives, and this alone should have given you reason to pause, but evidently you found it easier to assume I was stupid or lying. Oh well.

Given that nothing you say here is new, but the errors with the assumptions you make again have been pointed out repeatedly, perhaps you really need to start paying attention yourself.

Let me give you a clue that you will need to answer (either to yourself or here) before your answer becomes relevant:

Why is the distinction THE ONLY WAY to do what you think it is doing?

And yes, I do know what you think it is doing, and I also know that it is only your sheer arrogance and belief in your own self-righteousness that this is the ONLY WAY.

In your response, please spare us the concomitant condescension and "desire for compromise" pretense. It doesn't wash, and never has. Just answer the question, and keep the fluff to yourself.

Thanks. -- Fullstop 04:35, 3 December 2007 (UTC)

Both questions were answered directly. No one said this was the only way to address the issue. And with that, I'm done discussing this matter with you until you can join in with some minimum level of civility and maturity. Vassyana 09:30, 3 December 2007 (UTC)

Let me ask a question here. Suppose we used language like this: "Because the notability of a subject cannot generally be determined by the expert who first invents, discovers, or presents it, notability claims need to be sourced to experts with a reputation for reliability for evaluating notability in a field, and who are not directly connected to the initial invention or discovery. However, once notability is established by a neutral expert, the initial discoverer can be used as a reliable source for other facts about the discovery." A crude first start, but consider it. It seems to me that language like this could potentially describe the entire problem in terms of core Wikipedia terminonlogy like "notability", without the need to introduce new and potentially confusing terminology like "primary", "secondary", etc. It seems to me nobody really cares what level any given source is. All we care about is whether someone has evaluated a claim independently of the person presenting the claim, and when the presenter is recognized as an expert we only care about this for certain purposes like notability, not for all purposes. Why not say this directly? Wouldn't it make things simpler? Why the extra terminology and classification baggage? Why not stop by Occam's shop and try out a shave just to see how it would feel? No need to commit -- we can always grow the jargon thicket back if we find we can't work without it. Best, --Shirahadasha 06:08, 3 December 2007 (UTC)

Thomas Kuhn says it better. When there's enough of an accumulation of things that don't quite fit, it time to start rethinking. Rethinking starts by special casews and add-ons that end up making things more complicated; when enough people find this enough of a burden, when things start feeling kinda scraggly, it's time to start looking for simplifications and to start thinking about a shift and a shave. Perhaps we're somewhere near that point. Best, --Shirahadasha 06:22, 3 December 2007 (UTC)

Unfortunately, there is enough justified opposition towards using other terms to prevent the change. I've advocated for a shift in terminology in the past. Vassyana 09:42, 3 December 2007 (UTC)

(outdent response to Vassyana's earlier comment)

If you had answered my original question in a rational fashion (instead of trying the "to be very blunt" non-sequiturs again), I would have reiterated the problems that your premise has. Instead you were being provocative (and do so again). As I've told you before: don't dish out what you won't eat yourself.

Anyway,...

1. The focus on primary sources being more susceptible to OR than secondary sources is a chimera, a clustering illusion. That one group of sources are more prone to misuse can just as well be because they are more commonly available. It is the numerical predominance of primary sources that can make it appear as if they were the more commonly abused kind of source.
2. Inversely, secondary sources might appear to be less prone to abuse because traditionally the people who had access to secondary sources typically also knew how to use sources at all. (see also #7 below).
3. The P/S distinction is redundant when "we are allowed under policy to simply say The Gallic Wars reads 'X'."
   "Simply say [Y reads] X" - a.k.a "stick to the source" - is the fundamental tenet of NOR, and it doesn't then matter whether the Gallic Wars are primary or secondary. "Simply say X" applies to *all* sources.
4. "The necessity of the distinction arose due to the abuse of primary sources for original research" is based on the premise that the abuse was actually due to the "primary-ness" of the sources. I.e., the abuse was the fault of the abused, rather than of the abuser. Then, instead of tightening the rules to prevent abuse at all, what we got was a disqualification of the abused because they were abusable. Is this whacky or what?
   Said the judge to the abuser, "I see you beat your wife. I hereby throw your wife in jail. Then you won't be able to beat her."
5. The use of primary sources can be restricted without any need to categorize them. This can easily be accomplished by directly addressing what it is that makes primary sources primary. e.g. "the use of Gallic Wars is not acceptable as a source of history because history is what others make of an event." Or "because of the lack of perspective, any analysis by Caesar is not permissible."
   Allowing Gallic Wars (as far as NOR is concerned) can be expressed as: "the disqualification of de Bello Gallico as a source of history doesn't automatically preclude the policy-conform use of the book in an article on the book itself."
6. Perhaps the most frequent kind of OR is the - otherwise accurate - use of sources outside the context in which they were written. This kind of OR (e.g. citing Caesar's views on the Celts in an article on the Celts, George Bush's views on democracy in the article on democracy) won't go away by stemming primary sources when we also have secondary sources that reiterate them (in which case we might even get OR that is twice out-of-context).
7. OR abuse is rampant for sources that are accessible, and as more and more books appear on Google Books and as people begin to lock into it, the more often OR occurs for secondary sources as well. Most of the OR I see today from established editors is this kind of OR, and because NOR's PSTS addresses the symptom - but not the problem, which is abuse of sources in general - NOR is outdating itself. The infection today is not the infection of two years ago.
8. The fundamental problem of the PSTS section to NOR is the fact that the PSTS section exists at all. As explained above, the distinction between P/S/T has little or no real impact when *any* source can be misused for OR. But by focusing on what is effectively inconsequential, PSTS's distinction then actually also has a negative effect: it is a distraction.

-- Fullstop 20:34, 3 December 2007 (UTC)

1. Regardless of the reason, primary sources were more abused. It could be a matter of accessible sources, but it has not been established as the root cause and idle speculation doesn't do much for the discussion. (Also see 7.)
2. Libraries, online services and myriad other sources provide access to secondary material. This fairly universal access to such references in the Western world makes your claim dubious.
3. "Simply say X" =/= "Simply say Y reads X". This ignores the fundamental point I raised about certain sources being difficult to use without OR, except for "Simply say Y reads X"
4. False analogy, unless somehow any man couldn't be around that woman without beating her in most cases. As per 3, this ignores a fundamental point about the inherent OR problems in using some sources.
5. I agree with the essential point you raise (addressing what makes primary sources primary), but not with your elaboration. Other ancient histories are what "others made of history" (fulfilling your definition of history), but are still primary sources with the same problems in relation to OR as the Gallic War.
6. We can agree that out-of-context use is a problem, though we disagree on the degree and scope. (I've certainly advocated for language that insists on topical sources and in-context usage.) However, this does not alone account for problems with primary sources. (See 3 and 4.)
7. Feel free to prove this point and generate a consensus. Idle speculation as to the reason doesn't add much to the discussion. For example, one could argue this is the normal social process of displacement (over-simply, shunting behaviors from explicitly prohibited activity to "legally undefined" activity). Also, this does not equalize the nature and abuse potential of varying types of sources. (See 3, 4 and 6.)
8. You're certainly entitled to your opinion.

--Vassyana 23:45, 3 December 2007 (UTC)

1. No, not "regardless of the reason."
   • If primary source usage outnumbers secondary source usage by 4:1, and 10% of each are abused, then for 100 sources cited, there will be 8 primary source abuses and 2 secondary source abuses. Numerically, primary sources will seem to be abused more often than secondary sources, but in reality they are not.
   • You are obliged to have a reason. Otherwise...
     "Idle speculation doesn't do much for the discussion" includes the speculation that "primary sources were more abused," which is indeed a speculation that has not done much for the several megabytes worth of discussion.
2. So what if "libraries, online services and myriad other sources provide access to secondary material"?
   If people actually used them then we wouldn't have the preponderance of primary sources would we?
3. I didn't imply anything by putting the [read as] in brackets other than to say that "Simply say Y reads X" is a command to "stick to the sources." It does not "ignore" anything. See also #6 below.
   My original point #3 (that "the P/S distinction is redundant when 'we are allowed under policy to simply say The Gallic Wars reads X'") remains as valid as before.
4. If that is a false analogy for the reason you provide, then you threw your sole argument (that PS are susceptible to abuse) out of the window. "[As if] any man couldn't be around that woman without beating her in most cases" == as if any editor couldn't be around a source without misusing it in most cases.
5. The "elaboration" constitute examples, and its irrelevant whether you agree/disagree with the way they are phrased or not. The point was (and is) that "use of primary sources can be restricted without any need to categorize them."
   To reiterate: The use of certain kinds of sources can be inhibited even without giving them labels. They just need to be properly contextualized in relation to NOR, and not in relation to the kind of source they are.
6. The point was that OR won't go away by stemming primary sources when we also have secondary sources that reiterate the same thing that primary sources say. Even a direct quotation of sources can be OR (e.g. by taking it is out-of-context).
7. Yes, one could argue that increasing GBooks usage is the normal social process of displacement. But such a discussion would be ignoring the point I made and that you sidestep, which is that prohibiting the use of one kind of source does not solve the problem. The problem being OR, not "primary sources."
8. I'm sure that you are convinced that primary sources are not sources and thus require some kind of special treatment that sources are not generally subject to. But until I see some evidence to support it, you'll have to excuse me for considering your belief to be as useful as one in the easter bunny.
   True or not (and even if the definition of "primary sources" had not actually been made to order), it is still irrational to insist on a PSTS section that hinges solely on one (historical!) "primary sources" supposition.

-- Fullstop 03:46, 4 December 2007 (UTC)

1. A trend of misuse has been observed. Your idle speculation is no more accurate or useful than my own as to why. Observation of a trend and idle speculation as to its root cause are not equitable.
2. People being too lazy to use available resources is not the same as your original assertion that they lacked access.
3. Oh please, you explicitly equated "Simply say Y reads X" with ""Simply say X" (the latter is what you actually paired with "stick to the sources"). They are not the same thing. You are indeed ignoring my point.
4. Once again, ignoring the point. Some sources simply cannot be used directly (outside of pure "X says Y" or quotations) without engaging in original research.
5. By so defining sources, or placing sources in context in that fashion, you are by default categorizing them. It's simply a question of whether or not those categories are labeled and what categories you intend to use.
6. Once again, ignores the point that some sources cannot be used for standard direct citation without committing original research. I fully accept that out-of-context usage can be OR. However, that doesn't change or alter the point about primary sources.
7. I don't sidestep the issue at all. I've repeatedly acknowledged that secondary sources can be misused as well. However, here you once again ignore the point.
8. That's a severe twisting of my statements, to say the least.

Your twisting of other people's positions, brushing aside of central points and overall illogical approach (see for example, categorization) are tiresome. (I am not the only one to notice these traits, so it's not just me.) If you can productively participate, please do so. Otherwise, I see no reason to feed into your apparent need to turn this into debate club. Vassyana (talk) 23:10, 4 December 2007 (UTC)

Relation of PSTS to "... in popular culture" sections

Hi. I was directed here by an editor who disagreed with my removal of an "...in popular culture" section as "original research". I've read the policy here, and looked over the debate, and I'm confused.

My question is a general one, as this is not the only article in which I've seen (and sometimes removed such material. Nonetheless, for context I'll mention that the article in question was Wedgie, which is currently nominated for deletion. To clarify in general, I'm talking about a section in an article on topic X that enumerates instances in which X has been mentioned or shown in a television show, movie, video game, book, song, etc. My position is that such sections constitute original research because we're the first source commenting on that material, in connection with topic X. The sections consist of items of popular culture that Wikipedians have personally witnessed, noticed something about, and chose to document what they saw. That seems to me to fall well within a common-sense understanding of "original research".

In reading the debate above, I can't see that it applies particularly well to the type of situation I'm describing. The situation is not analogous to commenting that Einstein used five sources in his paper (the example given above); it's analogous to including a list of Einstein jokes that have been made in cartoons and TV shows. Is this question part of this debate, or am I barking up the wrong tree by posting here? What do people think about "...in popular culture" sections? Are they (potentially) original research? Is there another guideline or policy I should be considering? -GTBacchus^(talk) 01:30, 1 December 2007 (UTC)

Given that I hate most "in popular culture" trivia sections, I wish we could get rid of them by saying "Original research" but, alas, I don't think NOR really applies in most cases. Oh, I suppose that it could apply in a few individual cases... but as a generality, no. In most cases the bit of trivia is a simple declarative statement, of the ... "topic X was referred to in TV show Y (cite to episode)" variety. That isn't OR. It is simple statement and citation. One of the appropriate uses of the primary sources (ie the TV show itself). Now if the article went on to (for example) discuss the impact upon topic X that occured due to being mentioned in this TV show, we would need an outside secondary source to avoid OR. Blueboar 02:25, 1 December 2007 (UTC)

But those sections are absurd. By including them, we're claiming that, every time the Simpsons make a Marx brothers reference, it's somehow a piece of scholarship on the Marx Brothers to mention it. These sections are sometimes well-done (Ride of the Valkyries does a great job), but most of them push the limits of WP:NOT indiscriminate. On what do we base the judgment that a mention on a cartoon buys a mention in an article? -GTBacchus^(talk) 03:12, 1 December 2007 (UTC)

I realize this is not exactly the appropriate forum to discuss this issue, but, since you brought it up, the abhorrence of popular culture sections borders on the obsessive. In the article Grosse Fuge, for example, the fact that Beethoven's work is quoted in Schnittke's third string quartet was deleted unthinkingly, because the title of the section in which that appeared was called "... in popular culture". I know this is an unpopular view, but I see nothing unencyclopedic about listing and discussing derivative works of an artist or a work of art. --Ravpapa 12:42, 1 December 2007 (UTC)

It does not really matter whether we like "in polular culture" sections or not, nor whether we think it is appropriate to include such sections in our articles... The question that was raised was whether "in popular culture" sections constitute an NOR violation. The answer is that, in most cases, they don't. It's that simple. Blueboar 15:01, 1 December 2007 (UTC)

How would you reply to my question: On what do we base the judgment that a mention on a cartoon buys a mention in an article? Is that true in all cases? When is it not true, and why? -GTBacchus^(talk) 22:02, 1 December 2007 (UTC)

Addressing the initiating comment:

• With respect to individual entries in a pop-culture section:

There is no relationship between NOR and individual statements in a pop-culture section. The entries in such sections are rarely (ever?) backed up by citation. The permissibility of individual entries is controllable with WP:V.

• With respect to pop-culture sections as a whole:

The relevance of these sections to the primary topic is rarely, if ever, a given. The assertion/implication of relationship is the OR. It is, alas, not enforceable with NOR.

• Resolve the problem of the "in pop-culture" lists by shoving them off into an article of their own (distinct from the original subject) and tagging them with Category:Pop-culture lists.

-- Fullstop 23:57, 1 December 2007 (UTC)

Your solution doesn't work in situations where the article is too short to justify a separate "in pop-culture" list article. The idea that "the permissibility of entries is controlled with WP:V is false. For one thing, WP:V is just WP:NOR in different clothing. For another thing, doing original research, and then claiming that any reader can verify it by doing the same research is not compliant with WP:V. -GTBacchus^(talk) 19:16, 2 December 2007 (UTC)

1. Then consolidate! e.g. "Representations of historical personae in pop-culture" or whatever.
2. Oh no. WP:V is not NOR is different clothing.
   Not really relevant to the question you are asking (see below), but...
   • An unsourced statement can be tossed on grounds of V, but not on NOR (because the checker wouldn't have any source to compare with).
   • Also, establishing "verifiability" can be done by others. For a sourced statement, such verification would be to determine whether the cited source is RS or the representation of it is OR.
   • And, in the reading/writing cycle, V occurs after something is written down, i.e. after the OR would have occurred.
3. Indeed, its not kosher to claim that something is verifiable by "doing the same research." But why do you mention this?
4. To go back to your question "[does] a mention in a cartoon buys a mention in an article?"...
   No. A mention in a cartoon does not buy a mention in an article.
   BUT as long as a source is not provided, it cannot be rejected on OR grounds. It can only be rejected for not being V.
   Then, when the source has been provided, it can be rejected on OR grounds (i.e. for the contextual reasons).

-- Fullstop 23:15, 2 December 2007 (UTC)

Hi, thanks for your reply. Point by point:

1. I don't see how this suggestion applies to the kind of problems I'm talking about. I'm talking about long lists of every time, in some cartoon, someone makes a Freud joke, adding that to our article on Sigmund Freud. How does "consolidating" solve this problem? Even worse, how does it help an article such as Wedgie, which consists almost entirely of a list of pop-culture references?
2. I suspect you misunderstand what I meant by WP:V being WP:NOR in different clothing. The three core content policies of WP:NOR, WP:V and WP:NPOV are deeply complementary, and can be seen as three faces of the same general idea. They have the same spirit. How do you avoid doing original research? You only add what's verifiable. How do we keep articles neutral? We don't do original research, and we restrict ourselves to assertions that can be verified. How are assertions verified? Not by original research, but in reliable sources that we can attribute in keeping with NPOV. There's a singular principle uniting all three within the context of Wikipedia.
3. I mention "doing the same research", because that's the claim people have been making. Consider an entirely made up example: User:Prima watches a TV show, notes that they made a reference to Quasimodo, and then adds that information to the Quasimodo article. User:Secunda removes it (per WP:V or per WP:NOR; the way I'm thinking of it, it really doesn't matter). Prima defends it by saying that it's not original research, because the TV show is a source, and that it's verifiable because you can watch the same TV show, and see that they did indeed reference Quasimodo.

   My contention is that Prima is in the wrong, because he is arguing that you can verify the information by doing the same research (watching the show) that he did. Does that clarify my point?

4. I don't see how it's very important whether a piece of information is rejected on grounds of V or NOR. The point is, it's not information that we can verify in a source, so it goes. When you say "when the source has been provided, it can be rejected on OR grounds (i.e. for the contextual reasons)," I don't understand what you mean. Does stating the specific episode in which a reference occurred count as verification? What does "contextual reasons" mean in connection with NOR? -GTBacchus^(talk) 23:35, 2 December 2007 (UTC)

Explaining...

1. "Then consolidate them!" refers to your observation that "[moving pop-culture sections to their own article] doesn't work in situations where the article is too short to justify a separate [] article."
   "Then consolidate them!" means ... put multiple pop-culture sections in one article, e.g. in "Representations of historical personae in pop-culture" or whatever.
2. >>>"How do you avoid doing original research? You only add what's verifiable."
   NO! You avoid original research by sticking to the sources!
   Anything you write is inherently verifiable when you stick to the sources.
3. Yes, I understood you the first time already. But it doesn't matter. Anyone claiming that something is verifiable by "doing the same research" is violating V. End of story. All it should take is a tough admin to enforce this, but unfortunately we don't have very many of those.
   • >>> The point is, it's not information that we can verify in a source, so it goes.
     a)a WP:V violation is an unsourced/unsourcable statement. It cannot be verified because it does not have any source to verify it against.
     b) a WP:NOR violation is a sourced statement that cannot be verified against the source that is being cited.
     a) Statements in a pop-culture section that are not citing sources can be tossed for V.
     b) Statements in a pop-culture section that are citing sources can be tossed for OR (for not reflecting context, see next bullet point).
   • >>>> [I don't understand] "when the source has been provided, it can be rejected on OR grounds (i.e. for the contextual reasons)"
     OR is the act of not precisely reflecting what the cited source said, which includes the context in which the cited source said it.
     A Marx brothers reference in the Simpsons ... is a Marx brothers reference in the Simpsons.

-- Fullstop 00:43, 3 December 2007 (UTC)

Ok, #1 is down to a misunderstanding. I didn't mean that the pop-culture references section is too short for the article, I meant the article without the pop culture section is too short. An article such as "Representations of historical personae in pop-culture" would be the longest one in Wikipedia. Those sections are too long, not too short.

I don't see the point of your "No" on point #2. I think we're saying the same thing. When I say "only add what's verifiable", I don't mean anything different from "stick to the sources", because by "verifiable", I mean "verifiable in sources". I'm pretty sure we agree on this point, except for the "No."

I don't follow most of what you said under point #3. Are you saying that a Simpsons reference to the Marx Brothers, referenced only by naming the episode, belongs in the Marx Brothers article, or not? -GTBacchus^(talk) 00:51, 3 December 2007 (UTC)

Ok,...

1. an article without a pop-culture section is still an article in its own right. The benchmark of whether a subject warrants an article is governed by WP:N, not by the number of pop-culture references to it.
2. means "do not confuse V and NOR." If you will, you could consider the legitimacy of a WP statement being determinable in a series of three steps:
   a) Is the statement accompanied by a citation? If not, it fails V because its verifiability cannot be established at all.
   b) Is the citation accompanying the statement from a reliable source? If not, it fails RS.
   c) Does the statement accurately reflect what the source says? If not, it fails NOR.
3. What I am saying is ... A statement in WP fails the NOR test if it does not accurately represent the source that the WP statement is purportedly from. "Accurately represents" means that the WP statement may not convey something that is not explicitly evident in the source.
   When a Simpsons episode makes an allusion to the Marx brothers, the context is the Simpsons, not the Marx brothers. Since the Simpsons are not making a statement about the Marx brothers, the allusion cannot occur in an article about the Marx brothers.
   Note: Like any other violation of OR, such implied-relationship OR can occur with any source, even the most reliable kind.

-- Fullstop 02:05, 3 December 2007 (UTC)

1. Nothing further to say.
2. Yes, you're saying that it's important that I not confuse WP:V and WP:NOR. I'm saying that I prefer thinking about them as different aspects of the same general idea. Reasonable people may hold both of these positions; cheers.
3. Again, whatever. We're essentially agreeing, but you seem not to like the words with which I agree with you. I think you and I are both holding that these pop-culture sections consist mostly of original research, but we seem to be in the minority on that point. -GTBacchus^(talk) 03:10, 3 December 2007 (UTC)

Oh sure, we're in agreement. While its impossible to get rid of the unreferenced pop-culture entries on the basis of NOR, they can be gotten rid off with V. Once they're referenced you could (hypothetically) whack them as OR, but a fundamental weakness of NOR policy makes that presently unenforceable. -- Fullstop 05:00, 3 December 2007 (UTC)

But this is giving up on improving them. They are not necessarily lists, and they are not necessarily unsourcable. they can always be backed up by citation from at least the actual media (hence the relevance to accepted iprimary sources). They can for anything moderately important be backed up by reviews and discussions, though in nonconventional sources, such as forums (hence the relevance of discussions of web stuff as RSs), For the most important, they can be backed up by true citations to conventional books or articles about them. (But those trying to delete such material dont accept even that as relevant, and try to challenge each individual citation with a degree of intensity usually found only in ethnic conflicts; this tends to discourage people from even trying).

I think we want to maintain standards--but flexible standards, in accordance with the 21st century--and a comprehensive encyclopedia that acknowledges the way we in 2007 get information on the subjects that interest us. WP still has too much 20th century about it. DGG (talk) 05:57, 2 December 2007 (UTC)

I would dispute the implication that I (simply because I would delete "in pop culture" sections) don't accept secondary sources as relevant. If the mentions of all the times wedgies occur on TV shows could be sourced to any kind of secondary coverage, I'd be delighted. You shouldn't assume you know what I would accept as relevant.

That aside, there ought to be a way to take unsourced crap out of an article. I'll ask my question again: on what do we base the claim that a mention in a cartoon buys a mention in an article? How is this not original research, to assert that every instance of a particular concept in any TV show is significant and encyclopedic? Anyone? -GTBacchus^(talk) 19:16, 2 December 2007 (UTC)

All encyclopedias make decisions about what material to include in an article, and what material to omit. Wikipedia is an encyclopedia. Wikipedia editors make those kinds of decisions. This is not original research, this is source-based research. Remember, the policy isn't "no research", it's "no original research". --Gerry Ashton 19:57, 2 December 2007 (UTC)

Ok, I think I'm not articulating my point very well, then. I'm talking about material for which there is no source, except for the primary source material, which might or might not be available for verification.

If there were an newspaper article, or commentary from any reliable source, in which someone discusses subject X, and discusses its presence in popular culture, then we would be perfectly free to use that material, in any way we see fit. We are not, however, free to take that source as a jumping-off point, and go and do our own research on topic X, by reading/hearing/watching other media and collecting references to topic X. I'm in favor of source-based research, all the way. That's not what we're talking about here.

An example: Consider three works: The Odyssey, an epic poem by Homer; Ulysses, a novel by James Joyce, and How Can You Be in Two Places at Once When You're Not Anywhere at All, a radio play by the The Firesign Theatre. Each of these works has been written about, and we've got articles on all three. Joyce references Homer, and The Firesign Theater references Joyce.

Joyce's references to Homer have been noted, in detail, in published commentary. Due to that secondary source material, we are entirely justified mentioning Joyce's novel in the Odyssey aritcle, under "Derivative works".

The Firesign Theater references Joyce quite explicitly, quoting at length from the end of Ulysses. It's unmistakable, and we rightly note this fact in the How Can You Be... article. There, we're using the show as a primary source regarding itself. However, unless some commentators have noted The Firesign Theater's references to Joyce (which they may have, I don't know), we are not justified in talking about the radio play in the article about the novel. That would be using the radio play as primary source material for a subject different from itself, and that is original research. Indeed, there it is, sitting uncited in a section with a {{trivia}} template at the top.

Does this example clarify the point I'm getting at? -GTBacchus^(talk) 20:44, 2 December 2007 (UTC)

Hi GT, we're allowed to make edits based on primary sources alone so long as the edits are descriptive and not interpretive. This means we can say, "X appeared in cartoon Y," using cartoon Y as our primary source. We may do that even if no one else has seen fit to mention it (i.e. if there's no secondary source). It becomes OR when we say, "X's appearance in cartoon Y as a voluptuous and beautiful, yet maternal, figure may have been prompted by the death of the cartoonist's own mother, after which he embarked on a 10-year course of Jungian analysis" -- unless we have a secondary source who makes that point.

Having said that, I find the OR you often see in those sections harmless and quite interesting. SlimVirgin ^{(talk)(contribs)} 21:07, 2 December 2007 (UTC)

I often find them interesting too, Slim, and I regularly clean up such sections by organizing them chronologically, or something. I'm not trying to begin, or justify, some kind of deletion spree. I do, however, think that such documentation can be taken too far, and I'm interested in thinking about what that means, and where we would draw the line. Would you agree that there is a line, or is it always appropriate to include any external reference to subject X in our article on subject X? -GTBacchus^(talk) 23:12, 2 December 2007 (UTC)

I agree that a line can be drawn, though it's hard to say where. The rule of thumb I use is that I leave the edits if they're harmless and not too silly, and if there aren't lots of them. Occasionally, if there are too many in any given article, I go in and make a purge. SlimVirgin ^{(talk)(contribs)} 01:40, 3 December 2007 (UTC)

This is not really the right venue to discuss where such a line might be. This is where we discuss NOR... and it isn't an NOR violation to make a simple declarative statement about a TV show, book or movie and support it with a citation to that TV show, book or movie. It probably isn't even a NOR violation to make the statement without citation support (as the support is implied by the statement). May I suggest asking about this at the Village Pump or at WP:Trivia? Blueboar 23:40, 2 December 2007 (UTC)

Well, what I'm trying to argue - or at least consider aloud - is that it is, in fact, an NOR issue. That's how I've been thinking about it anyway, and I don't feel anybody has convincingly explained to me that it isn't. I'll go somewhere else, if you think this question would be more appropriate there, but I don't agree that it's not an NOR issue. -GTBacchus^(talk) 23:44, 2 December 2007 (UTC)

I think the main reason not to include every mention of a work in every other work that mentions it is that Wikipedia is not an indiscriminate collection of information. The value of an encyclopedia is to provide concise information to readers. Mentioning that X is mentioned in Y in an article about X is not "novel narrative or historical interpretation", as Jimmy Wales described it. We get to make judgments about what is important enough to include in Wikipedia, and what isn't. If someone can find a legitimate reading of the NOR policy that says we can't do that, we should continue to make those judgments and change the NOR policy to accommodate what we must do. --Gerry Ashton 00:00, 3 December 2007 (UTC)

Ok, I can see removing truly trivial stuff per WP:NOT#IINFO. It appears that my reading of NOR is not shared by many (any?) others, so I doubt we need to reword the page. It's not as if I actually read it. I do find it kind of silly that we can't remove unencyclopedic content from an article unless we utter the correct five-letter incantation while doing it. -GTBacchus^(talk) 00:05, 3 December 2007 (UTC)

The benchmark of whether a subject warrants an article is governed by WP:N, not really. The benchmarks are solidly grounded in WP:V and WP:NPOV. WP:N hopefully guides editors on how to assess material. ≈ jossi ≈ (talk) 02:15, 3 December 2007 (UTC)

That is neither the subject of discussion, nor is that correct unless you are intent on splitting hairs for the sake of splitting hairs. Fine, lets split hairs: "notability is an inclusion criterion based on encyclopedic suitability of a topic. The topic of an article should be notable, or 'worthy of notice'."

I'm sure you have some wikilawyer reason to suppose that that does not mean "whether a subject warrants an article is governed by WP:N," or if it does that you now have reason to run off to change WP:N too.

Why don't you try to be constructive for a change? Oh wait, that wouldn't be Wikipedia:Do not disrupt Wikipedia to illustrate a point#Gaming the system.

-- Fullstop 02:43, 3 December 2007 (UTC)

Sarcasm, as well as demonstrations of lack of good faith, will be thoroughly ignored. ≈ jossi ≈ (talk) 02:47, 3 December 2007 (UTC)

Sorry. Faulty premise. My bad. -- Fullstop 04:44, 3 December 2007 (UTC)

Secondary sources need not be verifiable by non-specialists

The present PSTS policy formulation says that statements from primary sources need to be verifiable by lay-people, but the same rule does not apply to statements from secondary sources. So in other words, in a higher mathematics article, where nobody but trained people would be able to verify, I am absolutely barred from citing the primary source for some theorem, even though I can cite (as I should be able to) a journal article that uses that theorem as part of the proof of a new theorem. The only problem is, I can't use that journal article to state what that new theorem is, because the article is a primary source for the new theorem. I can cite the proof, but not what is being proven. Is this sensible policy?

Going back to the Jacques Derrida example above, the present PSTS formulation puts us in a bit of a quandry. Suppose that Derrida is deconstructing Husserl. This is certainly a secondary source for Husserl, but also a primary source of Derrida's deconstruction. Whether or not I can cite Derrida hinges upon whether the source is primary or secondary. If Derrida is primary, I can't cite him because the citation cannot be verified by a non-specialist. If Derrida is secondary, I can cite him because secondary sources can be verifiable by experts, too. How do we decide whether Derrida's deconstruction of Husserl in this case is primary or secondary, and thus, whether or not it is prohibited? COGDEN 11:11, 3 December 2007 (UTC)

• Which part of "All sources should be used in a way that does not give rise to new analyses, syntheses or original conclusions that are not verifiable." do you find so difficult to understand? Cite the fact that sources make an analysis, don't make your own. .. dave souza, talk 13:49, 3 December 2007 (UTC)

Simply citing Derrida for what he says is not a "new analysis, synthesis, or original conclusion", because it's Derrida's analysis and conclusion, not the Wikipedia editor's. It's also verifiable to a reliable source: Of Grammatology by Jacques Derrida. It's just that this reliable source (like any source in this field) is not verifiable except by someone with significant training in critical studies. COGDEN 20:13, 3 December 2007 (UTC)

There are plenty of sources completely accessible (both in terms of acquisition and understanding) to the average educated person. There are even plenty of references that are accessible about the more difficult philosophers and philosophies such as Derrida, Hegel and phenomenology. Vassyana 23:08, 3 December 2007 (UTC)

There are not "plenty" of them. And citing them alone would violate WP:NPOV, particularly for Derrida and other 20th and 21st Century philosophers who railed against oversimplification, and whose words have a poetic element that is erased by the secondary sources. COGDEN 01:36, 4 December 2007 (UTC)

As someone who's studied philosophy, I can tell you there are plenty of them, a veritable mountain even. You could easily fill a bookshelf with books just explaining the meaning and nuance of Hegel alone, for example. There's practically a cottage industry within academic publishing around creating explanatory texts for complex philosophers and philosophies. The philosophers' criticisms of "trite simplification" applies even more to summary style encyclopedia articles (such as we produce in Wikipedia) than to textbooks and other summary sources, since by nature of the format they are even further reductions of the material. Additionally, since many textbooks and similar summary sources explicitly discuss the varying interpretations of philosophers/philosophies, such sources would help us adhere to NPOV, rather than being a danger to it. Vassyana 01:48, 4 December 2007 (UTC)

• I edit mostly technical mathematics articles. Here's the wikilawyer answer: For the purposes of Wikipedia, peer reviewed articles that present new mathematics theorems are considered secondary sources about those theorems. This is completely independent of any meaning of 'secondary source' in history. Please note that the language in the policy is carefully written so that only field notes and results of laboratory experiments are primary sources, not results of nonexperimental logical reasoning. I watch the policy quite closely in that regard...
     Now, I would be happy not to rely on the previous paragraph's wikilawyering. What matters in practice is whether there is consensus that the source is accurately represented, and consensus that the source is not given undue weight. When we discuss these things for a math article, we never do so in terms of primary and secondary sources. — Carl (CBM · talk) 13:56, 3 December 2007 (UTC)

Indeed. i have asked this before, and never gotten a wholly straightforward answer. Actually, nm - I'll break a new section off for it so as not to derail. Phil Sandifer 14:50, 3 December 2007 (UTC)

The straightforward answer, consistently with what Carl and Dave Souza just said, is that if the article is an exception to the rule, don't try to apply it. I don't see any equivalent mention of the apparent inapplicability of NPOV to this realm either. Shall we also go over and parse WP:NPOV and determine what flaws exist in that policy because they don't apply to an article that is inherently a straightforward technical matter? Here, if the "any reasonably educated person without specialist knowledge" clause doesn't apply, then arrive at a local consensus that this particular clause doesn't apply to this particular issue. As to Cogden's comment above ("The only problem is ... I can cite the proof, but not what is being proven." ), yes, if there's a debate about the validity of the content, a statement about what is being proven will require either (1) a straightforward recitation of what the primary source says is being proven, or (2) a derivative secondary source that verifies the statement of what is said to be proven byy the theorem. ... Kenosis 15:19, 3 December 2007 (UTC)

A "straightforward recitation" is not good enough, because the "accuracy and applicability" of that recitation has to be "easily verifiable" by a non-specialist. In effect, the rule says that you can't cite any primary source unless all readers can understand it. Only works written for lay audiences are citable, unless it's a secondary source, in which case, it can be written for the Ph.D-level reader. COGDEN 21:02, 3 December 2007 (UTC)

I am uncertain what number of articles are exceptions to the rule, but it seems like a rather frighteningly large number. Phil Sandifer 15:21, 3 December 2007 (UTC)

As is also the case with WP:NPOV, except it shouldn't be frightening to anyone because there are no sanctions involved for good-faith editing. None of these editing policies are airtight, and all of them have aspects that don't apply to everything on the wiki. Even the limits of WP:V must be negotiated by consensus. I mentioned earlier that there's no citation at all for Ottawa being the capital city of Canada in either article. Is that a violation of WP:V or an exception to it. If I went in and demanded a citation, the local consensus could readily decide "no, it's verifiable and doesn't require an inline citation". And so it goes across the whole wiki. If I come into the article on the unnamed hypothetical mathematical theorem mentioned above and say "hey, that's a violation of WP:NOR because it only cites to a primary source", the local consensus is free, so to speak, to tell me what level of hell to go to. In the same vein, if I put an "NPOV check" template on that same article because it only represents one side of the theorem, the local consensus is free to remove it because the notion that there's any other "point of view" to the theorem is simply irrelevant. ... Kenosis 15:39, 3 December 2007 (UTC)

You seem to be acknowledging now that many articles violate PSTS, and they should violate PSTS. This is not a good policy, and if that is the policy, that some violations of PSTS are okay, then we should either (1) specify what they are, or (2) change the rule to allow what is good and prevent what is bad. Simply saying that "all rules suck, so why bother to change this one?", does not solve the problem. If NOR is overinclusive or underinclusive, let's fix it. We can rebuild it. We have the technology.

As to the comparison with WP:NPOV, there's no comparison. There are no good technical violations of NPOV. If you can point to one, I'll point to a part of NPOV that should be fixed. COGDEN 20:22, 3 December 2007 (UTC)

Please desist from twisting my views in support of the legalistic, logical-positivist microparsing of this aspect of editorial policy. WP:PSTS has far fewer exceptions where it doesn't apply than does WP:NPOV, and also far fewer exceptions than a strict interpretation of WP:V. And, as has become increasingly clear, the complaints are in large part reflective of an apparent preference of some editors to do original research in WP rather than to avoid original research. ... Kenosis 21:25, 3 December 2007 (UTC)

Then, as I said, point to one. Where is a "good exception" to NPOV?

As to your claim that some editors oppose WP:PSTS because they want to do OR, who are they and how do you know their "preference"? Back up your claims. COGDEN 01:36, 4 December 2007 (UTC)

A "good exception" are your words. A de facto exception to the application of NPOV would be where an assertion by an editor that an article is not NPOV is by consensus deemed completely irrelevant or irrational because it simply doesn't apply, such as where the article is a simple list. NPOV may also be completely irrelevant in a straightforward recitation of a technical matter like a standard mathematical formula or type of mathematical formula. There's no relevant NPOV discussion in boolean algebra, or Euclidean geometry vs. non-Euclidean geometry. In these types of articles, the content is either relevant to the topic or it isn't relevant, and "neutral" has nothing to do with the discussion. With WP:V, there also are numerous de facto exceptions to the policy wiki-wide. WP:V becomes an issue only where the content is questioned as to its validity. Despite the technical requirement that if someone questions the content, a citation is required or it can be removed, I just gave an example of one of numerous instances where, as a matter of practice, the policy can be overridden by consensus of editors with the simple statement "no, it's verifiable, and no citation is needed for Ottawa being the capital of Canada", thus overriding my hypothetical removal of this content per my strict literal reading of WP:V. ... Kenosis 20:41, 4 December 2007 (UTC)

Are you saying that there are fields where NOR simply doesn't apply? If so, I'd like to know what they are. Are you saying that NOR is irrelevant to math, physics, and philosophy articles? I doubt that's your position.

You'll note that WP:V explicitly states the non-controversial claims need not be cited (to any source, primary or secondary). So it's actually not an "exception". It's the rule. If you can identify the "exceptions" to NOR, we can add them to the article and make them part of the rule, so that they are no longer exceptions that need to rely upon Ignore All Rules, which is a last resort, not a philosophy to live by. COGDEN 22:20, 4 December 2007 (UTC)

There you go again, attributing presumptions to things that were not part of my statements. ("Are you saying that there are fields NOR where simply doesn't apply?" and attempting to imply that "good exception" was my intended meaning, and so on with me and many other users-- but this is a whole 'nother cause for discussion, perhaps another time, perhaps in a thorough review of your own manner of shotgun-style pseudological criticism, perhaps never.) I was using the word "exceptions" to situations in which the policy analysis is not relevant, and but for your apparent and openly demonstrated stubborn determination to find things to attempt to pick holes in, I believe you're adequately smart to have known exactly what I intended to point out, but instead chose to twist it some other way. (That's right! I'm withdrawing AGF!) Therefore, other than to say that your manner and chosen direction of argument has become largely disingenuous in my opinion, I'll willingly leave you with the sense that you've made whatever point you felt you were trying to make here. Good day, or night, or whatever it is where you are. ... Kenosis (talk) 23:57, 4 December 2007 (UTC)

I wasn't trying to twist your words, and I apologize if you were offended. I was just intending to ask rhetorical questions, because I do not think good rules can have fixable unstated exceptions. COGDEN 23:41, 5 December 2007 (UTC)

Does anybody have concrete instances of what the PSTS distinction accomplishes?

So far there have been a pile of instances where people have pointed out that the PSTS distinction is unhelpful - Derrida, math articles, etc. What does it accomplish? That is, can people who support this distinction give some in-the-field examples where that distinction has been the policy that has governed an improvement to an article? Phil Sandifer 14:50, 3 December 2007 (UTC)

Examples that have been given many times are BLPs. Disallowing material based solely on primary sources stops editors from rooting around in courthouses to find divorce proceedings, banktruptcies, and petty feuds that no secondary source has decided is worth mentioning. In the area I work in a lot -- animal rights -- the distinction is used to stop primary sources being analysed solely by Wikipedians. For example, if I watch a video of alleged animal abuse, I'm not allowed to say what I see, except in very bare terms -- no analysis, no comparisons with other videos or other allegations against the same establishment, unless a secondary source writes about the video in those terms. The limitation on primary sources stops Wikipedians from waxing lyrical, in other words. SlimVirgin ^{(talk)(contribs)} 14:58, 3 December 2007 (UTC)

Would you agree the issue with court records is one of due weight rather than one of accuracy? — Carl (CBM · talk) 15:11, 3 December 2007 (UTC)

The former seems to me to be a particular (and non-special) case of the "Wikipedia is not a tabloid" rule. That is, given that we would probably remove information about divorce proceedings or a bankruptcy even if they had gotten press coverage as they're salacious details unrelated to the subject's notability, we should have no problem doing it for ones that are sourced to court documents on the same grounds. The latter is trickier, but I suspect NPOV adequately covers it - you remove the lyrical waxings with a comment to the effect of "undue weight being given to a POV that doesn't even seem to be published anywhere." Or, at least, that's what I've generally done in situations like that. So based on these, I remain unconvinced of the necessity of PSTS. Phil Sandifer 15:12, 3 December 2007 (UTC)

Right - I don't completely understand what SlimVirgin is saying about animal rights videos. As an extreme example, consider a video taken by an extremist organization after they illegally break into a lab. That video is certainly not a "reliable published source", so we shouldn't cite it for any interpretations about the conditions inside the lab. Even if we were to cite it, we would attribute the opinions to the extremist organization, because of due weight; but nobody is surprised that an extremist animal rights organization would think there is abuse. — Carl (CBM · talk) 15:16, 3 December 2007 (UTC)

Primariness or secondariness here has nothing to do with it. Court records can be just as secondary as tabloid journalism. For example, the attorney may play an audiotape of surveillance to the court, and what is said transcribed by the court reporter. That's a secondary source. The courthouse also has briefs, which are also secondary as to their subject matter. What Wikipedia is really against is unreliable sources--sources that have not been published. An unpublished transcript is unreliable. But it's not original research. COGDEN 21:09, 3 December 2007 (UTC)

This semantic definition of primary/secondary is not how the sources are commonly treated in scholarship (as has been repeatedly explained and proven to you) and certainly does not reflect the common understanding and usage in Wikipedia. Vassyana 23:01, 3 December 2007 (UTC)

It isn't? What about Henige, David (1986), "Primary Source by Primary Source? On the Role of Epidemics in New World Depopulation", Ethnohistory, 33 (3): 292–312, at 292, doi:10.2307/481816 ("[T]he term 'primary' inevitably carries a relative meaning insofar as it defines those pieces of information that stand in closest relationship to an event or process in the present state of our knowledge. Indeed, in most instances the very nature of a primary source tells us that it is actually derivative.…[H]istorians have no choice but to regard certain of the available sources as 'primary' since they are as near to truly original sources as they can now secure.")? Just to name one. Do you have evidence that Henige, or many other works that say the same thing in different words, do not reflect commonly-accepted scholarly definitions? COGDEN 01:59, 4 December 2007 (UTC)

We've been over this again and again and again. I've provided multiple sources (on multiple occasions) showing the lack of universal definition, including differing definitions both between and within fields. I really have no desire to rehash past conversations again. Suffice it to say, it's not the operative definition in Wikipedia (though it could be considered part of the operative meaning used here). Vassyana 02:05, 4 December 2007 (UTC)

I've seen different definitions, but never inconsistent definitions. There are "rough-and-ready" definitions for beginners, but the same people giving those defintions give fuller, more nuanced definitions when speaking in academic contexts. It's like when your Catholic school teacher defines Christianity as "belief in Jesus", while later in Catholic seminary, the same teacher may explain that yes this is true, but then there's the whole bit about the Trinity and stuff like how to classify Manicheanism and the Bahá'í Faith, etc. The first "children's definition" is valid, and good for some purposes within a limited range of thinking, but when difficult classification issues come up, it doesn't reflect real academic thinking on the issue. COGDEN 23:54, 5 December 2007 (UTC)

The distinction and the paragraph in this policy has been very useful in educating new contributors to fiction articles. It was one of the essentials around which my rewrite of WP:WAF centered, because the unfortunately-not-only-cliché "fan editor" who enthusiastically starts writing about his/her favourite fiction-related topic often has no experience in evaluating sources for the semantic levels of material they can be used to back up. NOR is a subtle thing, and most violations happen innocuosly because editors don't know about the distinction between primary and secondary sources. So, speaking from that corner of Wikipedia, I say this paragraph (potential improvements notwithstanding) is indeed very useful. I dorftrottel I talk I 18:00, December 3, 2007

Fair enough. Though literary interpretation seems to me different than other syntheses I'm hard pressed to explain why, and it may just be that I'm an English PhD student and so I care more about literary interpretation than other things. I wonder, though, if it might not be better to have a PSTS section in WAF that would better and more specifically serve the needs of that page. That said, it's been months since I've looked at WAF, and I have no idea what sort of state it's in. Phil Sandifer 18:19, 3 December 2007 (UTC)

I think the main issue I have with removing this from generally applicable policy and moving it into specialised guidelines is that there is an increasing number of "policy-literate" folks who know how to wikilawyer their way around following a guideline, because it's just a guideline, not policy. Please believe me, I wouldn't say this if I hadn't witnessed it many times over. A compromise I'd agree to could e.g. consist of a less restrictive wording here in NOR which still marks the distinction as generally relevant and refers to specialised MOS subguidelines (as far as they exist) for details on how to deal with the issue in various specific fields. I dorftrottel I talk I 19:12, December 3, 2007

WP:WAF is exactly the opposite policy as PSTS. WAF states that primary sources are required, and any secondary sources must be backed up with primary sources. I don't necessarily think that's a bad policy, but it severely undermines and contradicts PSTS. COGDEN 21:15, 3 December 2007 (UTC)

Um... having just looked at WP:WAF, I think you are misstating things a bit. Yes, any "In Universe" information (ie non-interpretive material about the piece of fiction such as its plot or characters) needs to be cited to the primary source (the fiction itself), but it clearly states that interpretive stuff needs to be sourced to secondary sources. The guideline seems to match what PSTS says, and even points directly to the PSTS section. Am I missing something? Blueboar 21:38, 3 December 2007 (UTC)

First, WAF is a MOS how-to subguideline, not policy. Next, "any secondary sources must be backed up with primary sources" — where within WAF do you read that? Also, your opinion that it "undermines and contradicts" PSTS tells me that you probably haven't spent too much time around e.g. typical Star Wars related articles. Many only use primary sources (at best), and most articles are more or less pure plot summaries. When I rewrote WAF, I attempted to properly introduce the notion of PSTS into the guideline for the first time. That's why I insist that PSTS is so important, especially to WAF: People try to constantly water down those guidelines in order to allow trivia sections, allow exclusive use of primary sources, etcpp. All in all Cogden, I appreciate your input, but I think you must not know what you're talking about. Any attempt to educate others about encyclopedic standards is a tedious battle, and I daresay I managed to do a somewhat decent job at WAF. Try and compare the current state of that guideline with that from before my rewrite. I dorftrottel I talk I 22:56, December 3, 2007

Here's what I'm talking about: "primary information describes information that can only be taken from primary sources.... secondary information describes external information taken from and preferably backed up with secondary sources.... Use as much secondary information as necessary and useful...not more...." I guess you're right that primary sources here do not have to be backed up by secondary sources. So that means primary sources are favored all the more. You cannot use secondary sources to say something about "primary information" (because such information "can only be taken from primary sources"). The primary source is mandatory, and no secondary sources allowed, except to provide a little background information, but not too much. COGDEN 01:59, 4 December 2007 (UTC)

You mean "not more and not less"? Did you check where "not more" links to? Also, I don't believe you're seriously suggesting that it is somehow possible to write any work of fiction without using that work of fiction as a source, do you? But this is getting off-topic here. How about presenting your concerns at WT:WAF instead? I dorftrottel I talk I 02:26, December 4, 2007

No, I absolutely agree that any article on a work should probably cite the work it discusses, rather than some fanzine-filtered secondary source "version" of that work. Primary sources are almost always more neutral and more accurate than secondary source (with the proviso that almost all secondary sources are also primary sources too, for their conclusions, interpretations, and syntheses). By the way, I made a couple edits to the WAF article to provide for the possibility that secondary information can be proven by primary sources. For example, you can verify an author's age (secondary information) based on a published interview of him (primary source). Plus, while fanzines are unreliable secondary sources as to original source's fiction, they may be good primary sources as to the fanzine or the fanfic (if the subject is notable). I think this is one policy area where the primary/secondary idea is useful, so long as the policy is stated accurately. COGDEN 03:06, 4 December 2007 (UTC)

What I see WP:WAF saying is this: To write about what the book says, cite the book; to write about what it means, cite independent sources. This seems perfectly sensible, but it certainly contradicts the idea that secondary sources are universally superior to primary sources as a basis for article content. Fortunately, we seem to have jettisoned that bit of excess baggage from PSTS, and now have a more balanced approach to the use of each. I still fail to see the necessity to engage in source typing at all. Sources must be used appropriately, and what is appropriate varies on a case-by-case basis. Dhaluza (talk) 00:56, 5 December 2007 (UTC)

Fiction in Wikipedia

Fiction on Wikipedia is one of the worst examples of laziness over good standards. The citing of the fictional material directly is most often justified by the supposed lack of sources addressing the topic. Like nearly all such complaints about a general lack of sources, this is just not true. (And regardless, if we don't have enough independent sources for an encyclopedic article, we shouldn't have the article.) For any given show, even short-lived and relatively unpopular programs, there is a mountain of periodicals that explain the show's principal characters, plot and so forth. Such publications come in both mass-market and industry-orientated varieties. For popular shows, the coverage is intense within those periodicals and also includes numerous books dedicated to the program. There are OR elements of interpreting character traits and plots involved in directly citing artistic works. There are additionally NPOV concerns regarding what elements of the story and characters are important to cover that cannot ever satisfied by directly citing the primary material. To be honest, I find this appeal to laziness to be one of the most pathetic and disheartening aspects of the Wikipedia community. We're supposed to make a high-quality encyclopedia, accommodating those who cannot or will not access quality sources is not the route to take. Vassyana 03:17, 4 December 2007 (UTC)

I agree... and what you say is also applicable to plots of movies. The editors at the Wikiproject Films, have agreed that it is OK to use movie itself to write a plot, in what could be considered a violation of NOR... ≈ jossi ≈ (talk) 03:21, 4 December 2007 (UTC)

As long as there is no interpretation going on, there is no violation. Making descriptive claims is not interpretation. If I say, "In his first mission as Agent 007, Bond goes to Madagascar in pursuit of an international bomb-maker named Mollaka," that's not original research, that's stating precisely what happened in the film, which is easily verifiable by watching the movie.  BIGNOLE  (Contact me) 03:29, 4 December 2007 (UTC)

I agree. Plot summaries, so long as nobody would disagree with them, are fully consistent with WP:V, and are not original research. Plus, WP:WAF is a policy. Now, you can argue that WP:WAF does not reflect true consensus, but then I'd argue right back that PSTS, too, does not reflect true consensus and should be changed. It's Policy vs. Policy. Who will win? COGDEN 22:29, 4 December 2007 (UTC)

WP:PSTS: "To the extent that an article or particular part of an article relies on a primary source, that part of the article should... only make descriptive claims about the information found in the primary source, the accuracy and applicability of which is easily verifiable by any reasonable, educated person without specialist knowledge, and make no analytic, synthetic, interpretive, explanatory, or evaluative claims about the information found in the primary source, unless such claims are verifiable from another source. Examples of primary sources include...artistic and fictional works such as poems, scripts, screenplays, novels, motion pictures, videos, and television programs." The only concern with writing plot summaries is if a writer strays from the limitation of being descriptive and crosses into interpretations and assumptions, which would then be a violation of WP:NPOV and WP:NOR. It's the same process as consolidating any source on Wikipedia -- we rewrite the content in a new manner, and we need to review it to ensure that the information stays descriptive. —Erik (talk • contrib) - 03:46, 4 December 2007 (UTC)

Thanks Erik, and Bignole ... the question Vassyana is asking: Why not to use secondary sources that must be abundant? Why to run the risk (which I have witness in several articles about films) of violating NPOV and NOR in such plot summaries? And what bout our readers? How can they trust these plots? ≈ jossi ≈ (talk) 03:54, 4 December 2007 (UTC)

(edit conflict)

The readers can't and they shouldn't. :)

Incidentally, only yesterday I fixed a case of obvious OR in the plot section of a Simpson's episode article that is both GA and candidate for FA! I hadn't seen the episode, but the OR was obvious anyway.

This OR wasn't caused by agenda or anything like that; it was just plain ignorance of how to guard against OR. -- Fullstop 04:08, 4 December 2007 (UTC)

Jossi, the same question can be asked about any source that's implemented. At the beginning of my editing run, I used only online sources. Then I learned to tap databases like Access World News. I'm citing something that some people may not be able to access right away, so I could easily introduce false information about a topic, such as citing reviews but re-wording them to make it sound like a critic liked it or disliked it. I think the key here is WP:AGF. I agree with Vassyana that a lot of fictional topics on Wikipedia need massive care. The problem is that fans write about what's near and dear to their heart, which may not fit WP:FICTION or WP:NOT#PLOT. I participate in quite a few AfDs to expunge in-universe compilations. Research is not easy for everyone -- a comic book fan would rather be wrapped up in the DC Universe and its superheroes, rather than looking at interviews about why writers and artists chose to take this particular approach with the storyline and the characters. There are users like TTN and Collectonian that are working very hard to compress fictional topics into real-world context, especially TV show articles that have nothing but plot detail in them. What needs to be encouraged are larger, more popular Wikias where fans can write everything about their favorite fictional topic, and they can come to Wikipedia for the real-world context. I've tried to suggest a transwiki to a Wikia (as one usually exists for a popular topic) in AfDs. —Erik (talk • contrib) - 04:01, 4 December 2007 (UTC)

Yes Erik, I understand your point, and appreciate your candor. These questions still remain: (a) What is better for Wikipedia: My summary of a plot, or the summary of a plot written in a secondary source; and (b) Why film plots are any different than an interpretation of the Bible? ≈ jossi ≈ (talk) 04:27, 4 December 2007 (UTC)

The summary of a plot written in a secondary source would still be rewritten by an editor if he or she did not quote at length. Whether we watch the film or have a script of the film handy, information is consolidated in the same manner. As for the Bible, I'm not sure what you're trying to ask here. Passages from the Bible are quoted without issue, but descriptions of copyright works like films or books are derivative works. We limit the content for this reason. I don't think there's truly any appropriate way to write out the plot of the film than watching it in many cases. The argument's been made that sources should be published, but this seems to me to be semantics. A film is in a static form and can be easily reviewed to present descriptive detail about the plot. Sorry if I'm all over the place here, 'cause I'm not quite clear what you're trying to ask. —Erik (talk • contrib) - 04:46, 4 December 2007 (UTC)

I believe our responses were directed specifically at your response. How can one trust any information cited on Wikipedia? How can I trust something you've cited from a book that I don't own, and isn't available online? How do I know you weren't just making it up, or using your own bias to interpret a certain passage a particular way? I don't. I have to trust that you are using the source appropriately, and I have to trust that someone else who has access to the source you used can verify that you weren't using your own original research. Online sources are not the only sources available, nor are they the best sources available. When someone cites anything that cannot be verified by simply clicking a link we are giving them our trust that they were not using original research, or violation a neutral point of view when they pulled from their sources. Citing a movie for itself is no different. I would suggest, that if there is any statements in a film plot that look "interpretative", that someone cite the minutes and the dialogue exchanged so that it can be readily available for verification by other editors.  BIGNOLE  (Contact me) 04:05, 4 December 2007 (UTC)

I wholeheartedly agree with Vassyana, especially with the last part: "We're supposed to make a high-quality encyclopedia, accommodating those who cannot or will not access quality sources is not the route to take." Clearly, not only must research for reliable, third-party sources be encouraged to back up plot summaries, but also the addition of real-world context material that requires secondary sources anyway. I think many shy away from adding anything but trivia and/or plot summaries simply because those require no additional research, or so they think. As for transwiki-ing unsuitable articles, Deckiller once started http://annex.wikia.com/wiki/Main_Page, but it wasn't a big success, for lack of centralised discussion. Strike that. The project appears to be alive, I just hadn't looked there for quite some time. Shame on me. I dorftrottel I talk I 04:20, December 4, 2007

Please show me where it says "no primary sources" in any policy. Not all information requires secondary sources. I can write up an entire production section of a film article with primary sourcing. I.E., interviews are considered primary sources, and I can write a production section on interviews with the cast and crew of a film. Secondly, you may not be able to find a film plot summarized well enough for "re-summarization"--yes, I mean precisely that, because we can't go steal a reviewer's personal description of the plot--for every film, even really notable ones. They may hit some key points they feel are relevant for their given review, but they may not hit ones that were relevant to the film itself.  BIGNOLE  (Contact me) 04:26, 4 December 2007 (UTC)

You're right, technically. The problem is that too many fiction-related articles are still a completely unreferenced mess. In the face of that, demanding research is a good thing. Even if a source cannot be copied, it would provide clues as to what the key aspects of the plot are and make for a very good reference in the article. And it would show that contributors are actually willing to research and read before they write. The problem is with accomodating those who are not willing or capable to do proper research (many are too lazy to even Google!), to read and use sources. Those few who are actually capable of writing a sound plot summary only using the work of fiction itself (and related primary sources) are incredibly outnumbered by those who can't. I dorftrottel I talk I 04:44, December 4, 2007

Read any number of other sources cited for other information, you could piece together what the film is about. Unfortunately, we are all outnumbered when it comes to people who cannot do proper research. You want to know what will happen if you pass a regulation that plot summaries must have secondary sources? One of two things. Either we'll have nothing but mass quotes from reviews, or people will stop writing them period. For some people, it isn't a matter of them being "lazy", but just not having the necessary skills to research well. You say "too lazy to even Google", but I know plenty of people who cannot even use Google to it's best efficiency. You have to remember that the majority of editors on Wikipedia aren't 35 year old PhDs, who have a background in research. I've met people from 12 to 40, with varying research skills between them, and not always in the direction you'd think. Just because we have editors who produce original research, or POV plot summaries doesn't mean that forcing them to find a secondary source will change any of that. What happens if I find a secondary source that isn't a url, or something you cannot view unless you have the physical copy in front of you? Are you going to say that the source is no good because you cannot click a link and verify the plot information I pulled from it?  BIGNOLE  (Contact me) 04:55, 4 December 2007 (UTC)

"just not having the necessary skills to research well"? If "lazy" constitutes a breach of good faith, saying someone lacks the necessary skills surely must be a personal attack... Anyway, people who can't even Google for a source shouldn't contribute to an encyclopedia project. I dorftrottel I talk I 05:20, December 4, 2007

No. You attacked their character, I was simply iterating that not everyone has the same skill level when it comes to researching; that was something you were basically alluding to when you stated that the reason the articles are in bad shape is because these editors are "lazy", and not because they may not know how to research very well. There's a difference. I know plenty of people who can research sources better than I can, but I know plenty that cannot do it at all, because they haven't developed the eye for good sources and bad sources. Not a huge deal, because it comes with time and practice. Regardless, you didn't answer my question. What makes it any different when you cannot view a literature source, and have to take the word of the editor citing that source? There is no difference. As easy as it would be to go down to the publich library, university library, or some other resource location, it's just as easy to get a copy of the movie/television show and verify the content stated, while at the same time making sure that they haven't introduced original research into the plot. Hell, you don't even have to do that, because you can usually spot potential original research in an article anyway, and if it looks suspicious, then just reword it/remove it/start a discussion on it on the talk page.  BIGNOLE  (Contact me) 13:00, 4 December 2007 (UTC)

Now, dorftrottel, you know that wasn't his intention. Simply put, some people are better and more skillful than others at researching material. What's wrong with speaking the truth? Liquidfinale ^{(Ţ) (Ç) (Ŵ)} 09:23, 4 December 2007 (UTC)

In addition, can someone tell me how citing a work of fiction itself for its plot summary is any different to citing a video interview with the director, a DVD commentary, or something else non-print, for (say) the production section? Well…it just isn't. It's not even different to my opening up a book or a magazine and paraphrasing for the same. Unless someone is willing to make a case for a Wikipedia-wide change to demand sourcing from written, online articles only, there's no argument to be had here, especially considering the problems we'd have in locating plot summaries for every article which requires one (especially older films, TV shows and novels, I would imagine). But I do agree that some plot summaries are either too long, too detailed, or contain too much interpretation ("Feeling conflicted over his twin desires to escape the city and to retrieve his father's watch…"), but instead of barring the use of primary sourcing for plot summaries, we should instead focus on properly educating Wikipedia users in the correct way of going about it. I would support, and participate in, a rework of the Plot guideline in the Manual of Style for Films (or elsewhere, if there's a better place for it) to get this across more forcefully. And not to hammer it home, but as long as the summary is written with as little interpretation as possible (e.g. "Butch hears a toilet flush; Vincent Vega exits the bathroom and Butch shoots him dead"), history has shown that there are relatively few problems beyond a sometime failure to adhere to the wider points of WP:NOR and WP:POV. After all, even WP:CITE says that attribution is only required "for...material that is challenged or likely to be challenged." Now, when there are problems with the method (e.g. "how do we know Butch hears the flush, eh?"), the routes open to us are, yes, find a citation for it if necessary. But also to look for a consensus; other people will have seen the film. The offending passage may even be removed if it has little impact ("Vincent Vega exits the bathroom and Butch shoots him dead.") You may not find this ideal, but the proof is in the pudding and, for the most part, it has worked thus far on Wikipedia. Difficulties emerge from time to time, but they're surely only a few moments thought away from resolution. Several times I've run up against a plot summary which contravenes all kinds of policies; it's been the work of ten minutes to rework some of them. And the problems (which I don't deny) of primary sourcing are still a drop in the ocean compared to the more common edit wars involving conflicting citations from two or more ostensibly-reliable secondary sources. Best regards, Liquidfinale ^{(Ţ) (Ç) (Ŵ)} 09:23, 4 December 2007 (UTC)

Thank you for a bit of fresh air in the midst of some excessive demands for a strictly legalistic interpretation of WP:NOR and WP:PSTS. Plot summaries do indeed get worked out by consensus, and their validity is verified by others who also are familiar with the particular work of fiction. PSTS isn't that complicated, and in the event there is a disagreement about the sourcing, it should indeed be worked out in a consensus process by those familiar with the work. According to PSTS, where a primary source is used as the source of a plot summary (e.g., the film or other original presentation of the work) the only restriction is that the plot summary should not make analytical claims about the plot. If someone puts in a statement in the WP article that, for example, "the final death scene of the film is a metaphor for the psychological death of losing a loved one [or, insert another preferred metaphor]", that will require secondary source(s) unless the film itself says "this scene here is a metaphor for the psychological death of losing a loved one [or other particular metaphor]". If people involved in the WP article want to use a written screenplay as a primary source, in addition to the film iself, they do so by consensus. If people involved in the article want to use secondary sources like commentaries and interviews to summarize the plot, there's no hint of anything in the policies to prevent this. The only proscription w.r.t. PSTS is against using the book or film itself to analyze the book or film beyond giving a straightforward plot summary that can be verified by anyone else who's read or seen the work. ... Kenosis 13:37, 4 December 2007 (UTC)

Thanks for all your comments. I would argue that this statement but instead of barring the use of primary sourcing for plot summaries, we should instead focus on properly educating Wikipedia users in the correct way of going about it, puts it quite well. The fact remains that many plot summaries are violating NOR, with too much detail, too much editorializing, and too much personal opinion, and indeed editors of these articles need to be cautioned not to cross the line into what is not permissible. ≈ jossi ≈ (talk) 16:25, 4 December 2007 (UTC)

But that has nothing to do with using primary sources, and all to do with simply informing new editors of the proper way to summarize a plot. Maybe the MOS for films needs to be a bit more specific. I don't know. But most new editors don't know any of the rules of Wikipedia, let alone the guidelines or style manuals that we have. It's less about what they are using the cite and more about how they just go about writing it in general. It wouldn't matter if you required secondary sources or not, because if you did, editors would still write plots without secondary sources. The only thing one would accomplish would be more needless edit wars.  BIGNOLE  (Contact me) 17:28, 4 December 2007 (UTC)

It has everything to do with not using primary sources such as the film itself to analyze the film itself. It is, though, completely consistent with policy to use either the primary source(s) or secondary sources, or both, as sources for the plot summary. Secondary sources such as reviews, and/or tertiary sources such as compendia of film plots, can be handy ways of summarizing plots and the policy actually encourages their use by way of stating that primary sources should be "used with care because it is easy to misuse them". But there's nothing in current editorial policy that prevents the use of primary sources such as the film itself for plot summaries, nor, IMO, should there be. If people want to bring articles on films or other works of fiction more into compliance with editorial policy, they should simply remove the extra commentary and analysis, along with any other "editorializing" about a plot, unless the analysis is drawn from secondary or tertiary sources. As a practical matter, such commentary probably should not be interwoven in a plot summary anyway. Like Liquidfinale pointed out, if someone cares to invest a bit of time to improve one or more of these articles, it appears fairly straightforward to remove most of the extraneous commentary with edit summaries like "removing personal POV from plot summary" or "Removing OR", or other appropriate description of the cause for the removal of the extraneous statements, adjectives, adverbs, etc. It sure would be nice to me if this weeding-out process were equally as striaghtforward in many of the Category:Philosophy articles. ... Kenosis 20:00, 4 December 2007 (UTC)

PSTS presently goes further than just barring the use of films to analyze the film itself. It also bars obvious and non-controversial summary or interpretation of the film—the type of summary that is explicitly allowed by WP:V and WP:WAF. The present PSTS language also bars using films to analyze other films. There is nothing wrong with citing one film that analyzes another film, such as This Film is Not Yet Rated. Films such as this are defined as a "primary source" for purposes of PSTS, even though they are a secondary source with respect to the movies analyzed. COGDEN 00:11, 6 December 2007 (UTC)

Just a comment releated to the above : remember that this should apply to all fiction on WP, not just films. I'm wondering if we should scope Writing About Fiction into MOSFICTION, of which then MOSFILMS can inherit from. --MASEM 18:14, 4 December 2007 (UTC)

law / articles about legal cases

I have some distinct opinions about the above (activist videos have recently shown up on another article, too; technical "source distinctions" should be used when helpful to explain things but should not, themselves, form the basis of policy or guidelines; I think of encyclopedia articles as akin to "review articles" in science) but that's all being said to death in one form or another. So here I have another area of writing & scholarship that hasn't been discussed but poses another set of issues; perhaps this will be a useful test case for people to test their opinions & proposals against.

Articles about law pose several issues for original research, particularly articles about legal cases. The opinion is itself a (cited) synthesis of research and is simultaneously the subject of other, conflicting opinions -- scholarly, popular, and legal/authoritative. Very few statements in a case unambiguously mean what they say -- we lawyers will certainly find ambiguity in any statement, over time. In practice, those of us who write on legal topics don't have much difficulty navigating this field; but attempting to apply the various iterations of WP:PSTS (I'm tempted to think of this as "WP:PTSD") clouds the issue, for me, at least. It's a good bit like the philosophy problem, above, except that citation practices are quite formalized in law and often very particular to individual quotes.

Discuss amongst yourselves. (-: Lquilter 16:44, 3 December 2007 (UTC)

My understanding is that court documents are considered primary sources. This would include the Judge's decision. Its perfectly acceptable to include a declarative statement about a legal decision in an article on something that directly relates to the case (for example: a biography of one of the litigants), and it is fine to cite that decision to back that declarative statement ... but any analysis or interpretation of the decision... any discussion of what that decision means beyond the narrow context of the individual case should be referenced to reliable secondary sources such as legal journals. Blueboar 21:54, 3 December 2007 (UTC)

Court documents are considered primary if they are a primary or original source of the information on the topic, but if not, then they are secondary. It all depends on how the court document is used. Any source "may be primary or secondary, depending on what the researcher is looking for". Monagahn, E.J. & Hartman, D.K. (2001), "Historical research in literacy", Reading Online 4 (11). COGDEN 02:20, 4 December 2007 (UTC)

By most scholarly definitions of "primary source", the Supreme Court is a primary source. It's not difficult to imagine a case where the Supreme Court disagrees with virtually all law professors. The Solomon Amendment case, Rumsfeld v. Forum for Academic and Institutional Rights, Inc., may have been such a case. How could we handle such an outcome? By most reasonable definitions of "primary source", the Supreme Court is a primary source. Do we really want to say that the U.S. Supreme Court, as a primary source, can't be quoted in a law article and that old law review articles must be preferred to its recent decisions when the two disagree? Functionally, the Supreme Court represents one of many examples of non-academic bodies that conduct peer review. Its function is to peer-review other legal decisions, and it only takes cases it considers notable (and its calls on notability are widely accepted in the field). Thus it functions in its field in a way that permits its decisions to be regarded as reviews of legal matters in a way that reliably permits basing notability on its actions. But by the primary-secondary-tertiary source schema, a Supreme Court opinion is a primary source. Best, --Shirahadasha 02:22, 4 December 2007 (UTC)

the decision is the the authoritative source for what it says, but not necessarily for what it means. The primary-secondary distinction is not relevant here, except to the extent that finding a reliable summary of even what the decision says may be best done from a secondary source--not necessarily even a law review, but a reliable quality newspaper. DGG (talk) 05:16, 4 December 2007 (UTC)

You absolutely may quote a Supreme Court decision... The only caveat is that you should stick to discussing exactly what the decision says, and not include any interpretation or analysis the decision unless you have a reliable secondary source that does so. Declarative statements about what a court document says (ie quotes from the document) can be cited to the primary source (the document itself)... Statements about what the document means (analysis or interpretation), on the other hand, need to be cited to secondary sources. Its that simple. Blueboar 14:32, 4 December 2007 (UTC)

This is true of any source. Why make a distinction between primary and secondary if the same rule applies? COGDEN 00:28, 6 December 2007 (UTC)

What about what a court says about another court? What if it's an inferior court's decision on a superior court's decision? ... Answering own question: I think you just have to say that "x-and-x courts have interpreted this as xyz, a view which has been followed by all the courts that have examined the issue" etc. I agree that secondary/tertiary is not helpful here -- "secondary" would be law reviews, newspapers, etc. -- but more relevant are the cascade of "authority": statute & cases first; followed by regulations & administrative interpretations & AG opinions; some of the most authoritative treatises or scholars that achieve widespread acclaim; all the rest of them; etc. This is what every first-year law student learns -- secondary/tertiary distinctions exist but are not that important, even in writing legal encyclopedias. --Lquilter 14:39, 4 December 2007 (UTC)

Actually, this is an important example of where secondary and tertiary literature are vital. Secondary literature is indeed inclusive of law reviews and newspapers, as well as numerous law summaries by competent lawyers and law professors. Tertiary sources include AmJur, legal hornbooks and a wealth of various other reliable legal summary material that specializes in integrating and analyzing the many important aspects of case law and statutory law. It most certainly is not our job in WP to be analyzing court decisions without benefit of secondary and tertiary sources. If the issue has to do with a specific court decision and how it affects public policy or other aspects of people's pursuits in the context of a particular WP article, secondary sources are similarly vital, and presently required by WP editorial policy so as to not conduct our own WP:Original research such as may be expected of law students, lawyers, professional legal researchers, law clerks and judges (indeed all of these frequently rely quite heavily on secondary and tertiary sources for their own background research). Any complaints within WP about reliability, or lack therof, of secondary and/or tertiary sources, can be discussed by participants in an article under a WP:Verifiability#Reliable_sources analysis. ... Kenosis 15:33, 4 December 2007 (UTC)

But why must we distinguish b/w secondary & tertiary? --Lquilter 16:25, 4 December 2007 (UTC)

There is absolutely no need to make any distinction between secondary and tertiary sources, let alone a hard-and-fast one. The utility of tertiary sources, though, wherever one chooses to draw the transitional stage between secondary and tertiary, is presently mentioned on the policy page. And the mention of tertiary sources also resolves the conceptual question of "what is Wikipedia?" as well as the question of what are other encyclopedias and other resources that feature summaries based upon very large numbers of primary and secondary sources. As with all the editorial policies including WP:V, it generally requires some degree of flexibility, some degree of respect for other people's often differing approaches and ideas, some degree of thinking about the editorial approach, some degree of interpersonal interaction, some degree of discussion as may be necessary, and some degree of reasonable judgment, in order to arrive at reasonable results that will be useful to a reader of the content. This is particularly the case where there are disagreements about a topic among those interested in that topic. Editors who are unwilling or unable to discuss with other editors how to arrive at a reasonable judgment in applying policy to editing practice, in my estimation, will tend to find editing very difficult with or without these core content polices. I say this in defense of the PSTS section in light of the obvious fact that the "lines" between primary, secondary and tertiary are not hard-and-fast and, as with other policies, may require discussion to arrive at a workable result in a particular wikiproject or a particular article. ... Kenosis 19:34, 4 December 2007 (UTC)

Well argued, thanks. A breath of fresh air... ≈ jossi ≈ (talk) 21:55, 4 December 2007 (UTC)

Don't breath too much of it. We still haven't made any headway in determining which legal sources are primary and which are secondary, and why it even matters. You apply the same rules for any type of source, regardless of primariness, because all legal sources are both primary and secondary, depending on how they are used. For cases, you might say that the true primary sources could be the briefs, court papers, transcripts, and precedent. After all, the judge is just commenting on the evidence placed before her—just creating a secondary source from primary materials. Every legal source is both primary and secondary, and there's no principled way to distinguish them except by pulling definitions out of our asses that have nothing to do with the idea of "primariness" or "secondariness" and add WP:CREEP. We need a principled policy. COGDEN 00:28, 6 December 2007 (UTC)

I'm sorry, but no. Any randomly chosen group of twelve lawyers would be almost certain to arrive at a consensus that legal cases and legislated laws are primary sources, IMO. And yes, a group of participants in an article about a legal topic could readily decide that so too are the briefs and testimony and other specific documents that make up the content of a single court case. And similarly it goes with legislation, where various documents are involved prior to the publication of the official legislation. I could understand that there might be some debate about whether, say, CFR is primary or secondary, but this kind of topic-specific determination should IMO be left to the local consensus in such a way that if a particular dispute requires WP:Comment, WP:Mediation or WP:Arbitration, that a group of reasonably educated commentators, mediators, or arbitrators can handle the situation as may be needed, without needing specialist knowledge to do it. ... Kenosis (talk) 04:47, 6 December 2007 (UTC)

In other words, without the ability of parties in an RfC, RfM or RfAr to point to secondary and tertiary sources that support an alleged primary-source interpretation, a highly technical topic will tend to seem like gobbledygook to the non-specialist, but with WP:PSTS there is at least a rational policy-based method within which WP users can point to other sources that back up the particular interpretaion at issue in a particular topic, without necessitating that only users with specialized knowledge will participate in such a content dispute where there is disagreement about it. (If there's no disagreement about it among the experts, there's generally no issue related to any policy at all, including of course WP:PSTS.) Does this solve every possible permutation of how such a debate could go? No, of course not. But IMO it goes a long way towards allowing WP to handle most of the reasonably forseeable editorial conflicts without amounting to excessive policy creep. ... Kenosis (talk) 05:21, 6 December 2007 (UTC)

In response to Lquitler's question above, I think the problem lies in what is the intent of a Wikipedia article on a legal topic? On one hand, many laws have effects on society far beyond the law books. An example of this I have in mind is a perennial crank argument that the Sixteenth Amendment is unconstitutional because of technicalities whether Ohio (one of the states that ratified the amendment) is legally a state. (I find this point entertaining, & hope that a history of this belief would not consitute orignial research & may never appear in Wikiepdia.) On the other hand -- & far more importantly -- is the article's intent to provide legal advice? In that case, sure it's OR but that is trumped by (wait for it) Wikipedia does not offer legal advice. -- llywrch (talk) 22:18, 4 December 2007 (UTC)

Discussion about image use and WP:NOR

Please come participate in the discussion here. It involves image use policy issues far beyond the template itself. Thanks. ···日本穣^{? · Talk to Nihonjoe} 06:51, 4 December 2007 (UTC)

COGDEN's tag

Let's discuss COGDEN's disputed policy section tag that was apparently created specifically for this dispute and has since been rediercted to a disputed policy tag, which is completely inappropriate. I think Mikka said it best about this tag, I'm searching for that quote now, but essentially if a section of Policy is disputed then it should either be removed or left until consensus is reached for it's removal. Dreadstar † 22:18, 4 December 2007 (UTC)

I do not care what tag is on this section, but there is a substantial dispute going on about it right now, and it seems honest to mark that dispute. I really don't give a fuck whether we mark it with an old template, a new template, or what, but marking it is appropriate. Phil Sandifer (talk) 22:21, 4 December 2007 (UTC)

It's really quite disruptive to keep on adding a tag. All policies and guidelines have disputed sentences or sections, and this particular section has been here for a long time. Disagreements should be discussed here, but not allowed to destabilize the policy. Cogden, I for one don't even understand a lot of what you're arguing, so please find a way to clarify what you're saying. Also, it would be helpful if you'd provide academic sources for your definitions, as it's not clear whether you're inventing them or taking them from another source. SlimVirgin ^{(talk)(contribs)} 22:23, 4 December 2007 (UTC)

Slim, his definition is drawn from an appropriate source (at least the most common one he uses). However, it's not even a universal definition within its field of origin. The problem with an "academic" definition is that there are several, varying both between and within fields, and that none of them match exactly with the Wikipedia usage (though some could be considered as comprising aspects of our operative definition). Vassyana (talk) 22:33, 4 December 2007 (UTC)

Policy changes should be discussed first, not made by one person. — Rlevse • Talk • 22:42, 4 December 2007 (UTC)

Dreadstar is correct about the tag template. Originally, it was {{disputedtag}}, but I changed it, creating a new tag, because there were some objections to it. The new tag was a compromise, but I agree now that we really should use the standard tag for these situations. COGDEN 22:57, 4 December 2007 (UTC)

May I suggest finding community consensus for any new or existing "dispute tags" before applying one to an Official Policy page. Personally, I don't believe such tags should be used at all on an official policy page - if the content is there, it's policy by consensus - disputed or not, if it's not there by any prior consensus and is disputed, then it should be removed - not tagged. Dreadstar †

Respondint SlimVirgin, the definitions I'm going by are from primary source and secondary source, and there are many academic citations there. COGDEN 22:57, 4 December 2007 (UTC)

Responding to Vassyana, I have not yet seen any reliable source stating that the definitions in the primary source and secondary source articles do not apply in some particular academic field. It's really the same set of definitions everywhere. Some web pages will lump various types of sources together, like diaries and maps, as primary sources, but in an academic article discussing what the terms mean, historiographers, historians, library scientists, and scientists all agree on what the terms mean, and that whether a source is primary or secondary really depends on how you use it. COGDEN 22:57, 4 December 2007 (UTC)

Responding to Rlevse, please read WP:BOLD and WP:EP, which I think are very good ideas. We've tried discussing proposals first then editing, but it never went everywhere because there is always some lone dissenter to any idea, no matter how good. For a while, being bold was getting us somewhere, but we've backslid a little. Maybe it's time to go into protection mode again, where we are forced to duke it out here. I'm open to all options. We can also try mediation again, but that didn't work the first time. This dispute has been going on since July, so I don't think anybody can say that the present version of PSTS represents widespread Wikipedia consensus. That's just not credible anymore. That's why we need the tag, because otherwise, people will see the tag at the top of the article that claims this represents widespread consensus, when it does not. COGDEN 22:57, 4 December 2007 (UTC)

When changes are as contentious as these have been, WP:BOLD doesn't really apply, Wikipedia:Consensus and Wikipedia:Resolving disputes do. A few editors here claim that this policy does not have widespread consensus, others maintain that it does and they have shown proof of consensus. A few disputing editors do not counter prior consensus, a new consensus needs to be found for the proposed changes. We've been round and round this issue too. Dreadstar †

Just a bit of history, the tagging issue has been going on since August, and there’s been no resolution to the issues it raises, nor the inclusion of such a tag on a Policy page such as the one Cogden proposes, this has been a hotly contested issue. The changes to OR that are still under dispute have been discussed since August, with some pointed commentary on the changes here, amongst other things about this ongoing dispute. We need to find a way to bring an end to this by either finding consensus for changes or leaving the policy as it was before the edit war started back in August. Dreadstar † 23:08, 4 December 2007 (UTC).

This edit summary by FeloniousMonk states part of the case for the removal of the tag perfectly. The tag has long outlived any justification for it, "you've had your chance to make your case. time for you accept that and move on". I suggest you make a solid proposal and put it up for consensus, if you like, but quit edit warring and tagging the section after all these months of getting nowhere. Dreadstar † 23:18, 4 December 2007 (UTC)

The anti-primary-source faction has had its chance to make its arguments over the last four months as well, and consensus has not yet been reached. But so what? Is there a time limit for establishing Consensus? If we can't establish consensus within four months, does controversy magically go away? This is simply a cop out for avoiding discussion. Reaching consensus is hard work, and there are no shortcuts or time limits. COGDEN 00:33, 6 December 2007 (UTC)

Frustrating

I find the current situation here frustrating. I'll acknowledge that no wording that is acceptable to everybody has been found, but there is a real dispute here, and it should have a tag.

I mean, I'd prefer to actually see some compromise on the wording so the tag can be removed. All I really want is something that loosens the wording up, acknowledges that these are general rules, etc. That's part of the problem - this is a policy page, and the applicability of PSTS seems more in line with the applicability of a guideline. (Or we could just spin PSTS off and tag it as guideline. That would even satisfy me.)

But really - can we actually make some progress here? A little? Maybe? And can we agree, in the name of good faith, to say that the section is disputed while we make said progress? Phil Sandifer (talk) 22:21, 4 December 2007 (UTC)

Can we not just add: "Like all advice and rules in Wikipedia policies, the decision as to whether primary or secondary sources are more suitable on any given occasion is a matter of common sense and good editiorial judgement?" SlimVirgin ^{(talk)(contribs)} 22:25, 4 December 2007 (UTC)

I would love it if we did. Phil Sandifer (talk) 22:27, 4 December 2007 (UTC)

Actually, lemme tinker the wording a bit, since this is a wiki: "Appropriate sourcing is a complicated issue, and these are general rules. The decision as to whether primary or secondary sources are more suitable on a specific occasion is a matter of common sense and good editorial judgment, and should be discussed on individual article talk pages." Phil Sandifer (talk) 22:29, 4 December 2007 (UTC)

That's fine with me. SlimVirgin ^{(talk)(contribs)} 22:30, 4 December 2007 (UTC)

The most destructive thing to consensus on this long-discussed section is an all-or-nothing attitude from people on both sides. (To simplify, generally meaning those who want to retain the section and those who wish to dismantle, or severely simplify, it.) This is not a problem exclusive to this policy. A number of rules and process discussions are derailed by such idiocy. So long as there is a solid group of entrenched people unwilling to seek consensus and compromise, and people are unwilling to ignore the obstructionists, no solution is going to be had. Vassyana (talk) 22:41, 4 December 2007 (UTC)

OK - what about the wording I propose above that SV likes? Any problems with adding it to the page? Phil Sandifer (talk) 22:51, 4 December 2007 (UTC)

It's good, but I see it as a Band-Aid over a gushing wound. Take a look at my proposal in the next section, which is a minimal change, yet at least from my perspective would move us 85% of the way toward reflecting actual Wikipedia consensus in real articles, be it in physics, math, philosophy, fiction, or whatever other kind of article. I don't see any loopholes here. COGDEN 23:07, 4 December 2007 (UTC)

I have no problem with such a caveat, at all. It would be a good addition. My comment was more towards echoing sympathy with your frustration. Imagine how those of who've been here for months discussing PSTS feel! Vassyana (talk) 23:23, 4 December 2007 (UTC)

Okay, I've added it. If there are no objections, I'd also like to remove the two sentences above it, which just repeat what's said elsewhere: "All sources should be used in a way that does not give rise to new analyses, syntheses or original conclusions that are not verifiable. Where interpretive claims, analysis, or synthetic claims are included in Wikipedia articles, use appropriate sources rather than original analysis by Wikipedia editors." SlimVirgin ^{(talk)(contribs)} 23:30, 4 December 2007 (UTC)

The first sentence is useful in making it clear that these requirements apply to secondary or tertiary sources as well as primary sources, to avoid the sort of tendentious misunderstandings so eloquently set out by cogden. The second sentence is well covered by the secondary source statement and by the new addition, and can be deleted imo. .. dave souza, talk 00:33, 5 December 2007 (UTC)

Proposal: just change the terms primary-->raw and secondary-->interpretive

After a bit of experimentation, here is a proposal that I could almost live with, and it's really a minimal one, but unfortunately it was reverted as kind of a knee-jerk reaction. Maybe that's part my fault for being to bold, but why would this not move us at least 85% of the way toward consensus?:

Sources may be divided into three basic categories of how they relate to the subject being written about. For the purposes of Wikipedia content policies and guidelines, raw, interpretive, and tertiary sources are broadly defined as follows:

• A raw source contains raw facts but no interpretation of those facts. Raw sources that have been published by a reliable source may be used in Wikipedia, but only with care, because it is easy to misuse them. For that reason, anyone—without specialist knowledge—who reads the raw source should be able to verify that the Wikipedia passage agrees with the raw source. Any interpretation of raw source material requires another reliable source for that interpretation. To the extent that an article or particular part of an article relies on a raw source, that part of the article should:

• only make descriptive claims about the information found in the raw source, the accuracy and applicability of which is easily verifiable by any reasonable, educated person without specialist knowledge, and
• make no analytic, synthetic, interpretive, explanatory, or evaluative claims about the information found in the raw source, unless such claims are verifiable from another source.

Examples of raw sources include archeological artifacts; photographs; historical documents such as diaries, census results, video or transcripts of surveillance, public hearings, trials, or interviews; tabulated results of surveys or questionnaires; written or recorded notes of laboratory and field research, experiments or observations, published experimental results by the person(s) actually involved in the research; original philosophical works, religious scripture, administrative documents, and artistic and fictional works such as poems, scripts, screenplays, novels, motion pictures, videos, and television programs.

• An interpretive source draws on raw sources to make generalizations or interpretive, analytical, or synthetic claims. Where interpretive claims, analysis, or synthetic claims about raw sources are included in Wikipedia articles, use interpretive sources rather than original analysis by Wikipedia editors.

• A tertiary source is a publication such as an encyclopedia or other compendium that sums up other sources. Many introductory textbooks may also be considered tertiary to the extent they sum up widely accepted results of large amounts of raw and interpretive sources. Tertiary sources can be useful in avoiding original research in topics where there exist very large amounts of raw and/or interpretive sources.

All sources should be used in a way that does not give rise to new analyses, syntheses or original conclusions that are not verifiable. Where interpretive claims, analysis, or synthetic claims are included in Wikipedia articles, use appropriate sources rather than original analysis by Wikipedia editors.

Frankly, I'd get rid of the "tertiary" section, but I want to make this the most minimal change possible at this point. COGDEN 23:04, 4 December 2007 (UTC)

I agree with loosing the tertiary section. It may be truth, but there is no need for it in a policy document. It belongs in Wikipedia. --SmokeyJoe (talk) 03:13, 5 December 2007 (UTC)

The problem with this is that you're substituting your own terms (raw and interpretive sources) for terms in common use (primary and secondary sources), and they're not equivalent anyway. A primary source can include interpretive material, for example. SlimVirgin ^{(talk)(contribs)} 23:24, 4 December 2007 (UTC)

(ec)This only serves to confound and mar the issue further. "Raw" sources are but a subsection of primary sources. For example, the Bible is not a raw source of facts, but it's certainly a primary source. I appreciate the good faith attempt to move forward. However, I don't think using language that either drastically narrows the scope of the definition or is contradictory to the operative definition is going to work. Vassyana (talk) 23:32, 4 December 2007 (UTC)

Raw sources are a subset of primary sources as I understand primary sources. But it seems that the rationale behind this language, back when it was written was directed toward preventing use of raw sources. What the policy really is trying to get at is the use of sources that do not contain within them published interpretive material. The original authors apparently didn't see any difference between raw and primary sources, and apparently thought that anything interpretive was a secondary source. So this is not really a departure from that original intent.

I don't see any philosophical need for us to stick with the terms "primary" and "secondary". The importance is the policy, not the terms we use. Go back and pretent that the PSTS section never existed, and you had never heard of the terms primary source and secondary source. How would you write the policy in your own words? COGDEN 23:07, 5 December 2007 (UTC)

Is there any real support or viable alernatives to change the terminology from “Primary, Secondary and Tertiary”? This has long been discussed, and I think we should try to lock it down now. Personally, I think we should keep the current terms. Dreadstar † 18:20, 5 December 2007 (UTC)

There's clearly no consensus for primary and secondary, but I think we can build consensus aroung other terms such as raw and interpretive. As I see it, nearly everybody opposing the primary-secondary language is receptive to some of the ideas in the PSTS section, but the main problem is that primary source and secondary source are ill-defined in such a way that it creates a whole host of problems. The terms carry too much baggage. If we want the policy to reflect consensus, which I think most of us do, I'm not sure we can get there with the terms primary source and secondary source. It seems like the simplest replacement terms are "raw" and "interpretive". Does anybody actually disagree substantively with the proposal as written above? Or at least not disagree. That could be a basis for consensus, and we could go from there. COGDEN 23:07, 5 December 2007 (UTC)

Please build that consensus for the actual terms quickly, this has been an ongoing issue for months and is continuing to drag things down. The current terms (Primary/Secondary/Tertiary) have clear and undeniable prior consensus, so a new consensus will need to be formed to replace them. Are there any other viable alternatives besides "raw" and "interpretive" (which I see have already been opposed above..an opposition which I join, as I prefer the original terms to any I've seen thus far). Let's lay them all out on the table.

As for the proposal above, I prefer how Vassyana has laid out the terms. Dreadstar † 23:16, 5 December 2007 (UTC)

For clarity

Is it correct that this is basically about the definition of the word "Primary source"?

--Kim Bruning (talk) 23:22, 4 December 2007 (UTC)

I wish it were that simple. :-P Vassyana (talk) 23:33, 4 December 2007 (UTC)

I'd recommend reviewing the following to get a better idea of the long-running discussion here:

• Wikipedia_talk:No_original_research/archive26#Alternative_proposals:_straw_poll.
• Wikipedia_talk:No_original_research/archive25#Goals
• Wikipedia_talk:No_original_research/archive24#Attempt_to_address_the_new_opposition
• Wikipedia_talk:No_original_research/archive24#Just_to_be_clear:-definition_of_primary_and_secondary_sources
• Wikipedia_talk:No_original_research/archive23#Back_to_the_question_of_what_policy_should_be
• Wikipedia_talk:No_original_research/archive23#Refocusing_on_the_source_discussion
• Wikipedia_talk:No_original_research/archive23#Why_we_DO_need_some_sort_of_a_statement_on_primary_sources_usage
• Wikipedia_talk:No_original_research/archive22#Before_we_unprotect_the_page
• Wikipedia_talk:No_original_research/archive22#Attention_newcomers_to_this_article

Vassyana (talk) 00:17, 5 December 2007 (UTC)

It's very complex, like Vassyana said, but biggest bullet-points as I see them are:

• The definition of primary source and secondary source (made up or standard academic);
• Whether those terms should be used in policy, or whether they are too high-falutin' and ambiguous;
• When, if ever, primary/raw/otherwise-bad sources should be prohibited/discouraged and when if ever, they should be mandatory/encouraged;
• Whether any of this belongs in No original research at all, or whether this is a Verifiability issue; and
• Whether the section needs to reflect current Consensus or whether it can remain unchanged by virtue of inertia based on apparent past consensus.

COGDEN 23:23, 5 December 2007 (UTC)

The issue of inertia and consensus for Policy pages was discussed in this section, which contains some compelling comments on this subject. I think Policy pages need to have strong Consensus to change prior consensus, and that policy pages need inertia to avoid chaos from constant changes becase an editor or small group of editors don't believe in the prior consensus - yet do not have consensus for their proposals, but they believe they can change or remove the material they dispute anyway. Dreadstar † 23:35, 5 December 2007 (UTC)

That's what a lot of people think, but that's not actually true, based on what the policy articles say. There is no "intertia" that keeps a non-consensus policy section active even though consensus in favor of that section is demonstrably lost. According to WP:POLICY, "Wikipedia polices may change as consensus changes, but policy and guideline pages must reflect the present consensus and practice." According to WP:CONS, "A small group of editors can reach a consensual decision, but when the article gains wider attention, others may then disagree. The original group should not block further change on grounds that they already have made a decision." Moreover, "In the case of policy pages a higher standard of participation and consensus is expected."

So really, the choice here is not between keeping PSTS vs. replacing PSTS, but between deleting PSTS and replacing PSTS. Either way, we have to end up with something with strong consensus. If that means "no consensus policy", that's what we'll have to accept, according to present policy. COGDEN 00:43, 6 December 2007 (UTC)

Again, you are mistaken. There is no present consensus for changes, so the prior consensus is what we go by. There is no reason to delete PSTS, there is strong support for that section now, all we're doing is attempting to rewrite it for clarity. The original group is not blocking your changes, the lack of consensus for your changes is. Accept that and move on. I'm not going to go round with you again, or repeat again the need for inertia in Policy articles, this is the same tune you've been playing for months on end. Move on and find consensus for your proposals. Dreadstar † 01:04, 6 December 2007 (UTC)

That is a novel concept of Consensus on WP. If that were true, consensus could be easily usurped by simply jealously defending an inaccurate transcription of actual consensus. By refusing to consent to change, all progress could then be indefinitely blocked. That is why the Consensus policy wisely allows consensus to change, and that policy must reflect current consensus, not a perception of past consensus. Naturally policy should not change on a whim, but it does change over time, and must be allowed to continue to change over time. Dhaluza (talk) 02:13, 6 December 2007 (UTC)

Revisiting a proposal

User:Vassyana/NOR 002 was suggested and repeatedly advertised on RfC and the village pump. It was one of the least controversial and most supported (including people on "both sides) of the various serious proposals to change PSTS. This suggests it may be worth revisiting it, even if to understand what worked "better" about it so we can create another draft. I'd like to review the comments made about it and see if we can adjust it sufficiently for use. Let's see what we can do to address any concerns, so we can move forward.

Support (some supporters included more than one):

• There was support based on the draft being a reasonable clarification.
• There was support based on the proposal better reflecting practice.
• There was support based on the draft being a reasonable compromise.
• There was support based on the proposal being a step in the correct direction.

Neutral:

• One person changed from opposition to neutrality after their concerns were addressed.

Opposition:

• There was opposition based on a strong opposition to PSTS, without feedback on the draft.
• There was opposition based on the presence of the source-typing caveat and the conflation of secondary and tertiary sources.
• One person opposed without explanation or feedback on the proposal.
• One person stated the status quo is clearer, without further explanation or feedback.
• One person expressed the draft is wordy and hard to follow, and objected to a requirement for explicit in-text attribution of primary sources.

Obviously, some positions cannot be accommodated due to a lack of feedback and/or explanation. However, I am going to endeavor to revise the draft accordingly to the reasonable opposition. The conflation of secondary and tertiary was founded in a concern that some summary sources (like textbooks) are exceptional sources and that this should be noted. The draft includes such language. There was less opposition to the caveat as a footnote, so I've moved it to such a format. I've moved a few other side comments and examples to footnotes as well, to help shorten/clarify the text. I've removed a chunk of text that may be better suited to another part of the policy. I've revised some of the language for clarity. (Additional comment) I also moved secondary before primary, because some concern was voiced about secondary sources being mentioned before their definition.

Regarding the requirement for explicit in-text attribution, I'm unsure of what other compromise I can make for this issue. Some expressed deep concern with primary sources being used in such a way at all, others argue that perfectly usable primary sources make such claims. The latter was accommodated with clear language permitting the use of such claims, but former accommodated with a requirement for blatant in-text attribution. If someone has a better compromise between the two extremes, it would be welcomed.

Please review the altered proposal. Let me know what you think. Vassyana (talk) 01:00, 5 December 2007 (UTC)

Thanks Vassyana. The lead of the section could benefit from better grammar and simpler wording, but overall I see this as a good summary that may be acceptable to all involved. ≈ jossi ≈ (talk) 03:05, 5 December 2007 (UTC)

Revised/simplified the lede language. Anything else that can be tweaked? Anything particularly flawed or lacking about it? Any additional concerns from the long long discussion here that you think could be accommodated? Vassyana (talk) 11:46, 5 December 2007 (UTC)

I don't see any significant problems. — Carl (CBM · talk) 13:31, 5 December 2007 (UTC)

To my mind, everything after (2) is still an overly strong version of the advice. Not wrong, as such, but overly restrictive, and without enough awareness that the rules do not apply in many circumstances. The wording SV proposed above about good judgment and the rewording of your suggestion below both open the door to the possibility of grey area in a way that the tail end of this doesn't.

That said, I love everything before (2), particularly the abandonment of the secondary/tertiary distinction. Phil Sandifer (talk) 14:50, 5 December 2007 (UTC)

I'll agree it is one of the more restrictive parts of the proposal, largely because it adheres closely to the current policy. I included language about explicit attribution for analytic et al claims from primary sources and about primary sources in secondary sources, because everyone can agree that both practices (as such) are OK. I will continue to think about how that portion can be improved and revised to provide an acceptable compromise for the policy editors and better reflect practice. However, I think any significant departure from that model/advice is really another step in and of itself (much like a change of terminology would be). Thanks for the feedback! Vassyana (talk) 17:19, 5 December 2007 (UTC)

I don't think we need a significant departure from that model... I think, actually, a big part of it is that that portion is best considered as advice and a guideline, unlike the rest of NOR - it's a small bit of not-quite-policy in a policy document. And thus some hedging to reflect that is in order. Phil Sandifer (talk) 18:08, 5 December 2007 (UTC)

I'm pondering this and if I think of a good solution, I'll share *chuckle*. If you have any suggestions about how to alter the wording, they would be welcome. Vassyana (talk) 03:26, 6 December 2007 (UTC)

This is an excellent proposal, Vassyana. I think it does a very good job of addressing a lot of the concerns expressed about the current version and it keeps the core of the policy intact. I like it. Dreadstar † 17:22, 5 December 2007 (UTC)

Yes, I like it too, particularly the way you have changed the emphasis of the section from the sources themselves to the purpose of the section, thereby better integrating it into the policy. I assume it would be under a heading of "Types of sources" or similar? A minor point, but perhaps the first sentence could be simplified to something like "To help identify and avoid original research, this section broadly defines primary and secondary sources." I found the reference to a tool unnecessary and slightly confusing: I couldn't tell if it is the section or the collective primary and secondary sources that is meant to be the tool.

Sorry for popping up here suddenly, by the way. You haven't seen me, but I've been working my way through this page and trying to follow its convoluted discussions for a couple of weeks. (Only a couple of weeks!) This is the first time I've felt able to contribute something. Hope I'm not making a total ass of myself!  —SMALLJIM  21:54, 5 December 2007 (UTC)

No apologies needed. I took your helpful suggestion for rewording. Are there are problematic phrasings? Any notable flaws (in general) in the draft? Vassyana (talk) 03:26, 6 December 2007 (UTC)

• Here are my comments on the proposal:

1. It's better than what we have now, and I support any change that moves in the right direction, with the proviso that it is not the final resting place.
2. Why keep the terms primary and secondary if we are now entirely divorced from the academic terms primary source and secondary source and from the concepts of "primariness" or "secondariness"? Why not combine my proposal with yours and use the terms raw and interpretive, and then just define them how we want? In fact, I notice you use the term "raw" in the definition of primary source. The beauty of using the term "raw source" is you can define it however you want, and don't have to worry about anomalies like peer-reviewed articles being secondary sources. They'll just be "interpretive sources", which makes perfect sense.
3. You are keeping the "reasonable, educated person" standard of super-verifiability for "primary" sources. I can see the benefit of an extra layer of verifiability for "raw" sources, but not in "interpretive" sources that are also primary sources, such as mathematical and philosophical works. We don't want editors interpreting raw data, but we do want them relying on technical primary sources like Principia Mathematica, Philosophiae Naturalis Principia Mathematica, and Of Grammatology.
4. The requirement for finding an explicit reference in the primary source for any interpretive statement contradicts WP:V, which does not require citation to any source when you make claims that nobody would dispute, such as "The book has 100 pages". You won't find that statement explicitly stated in the work, but so what? If you replaced "primary" with "raw", you would not even need this requirement because, by definition, raw sources have no interpretations to which you can cite, explicitly or otherwise.

COGDEN 02:05, 6 December 2007 (UTC)

I really think tackling terminology would be a whole proposal unto itself. I may not necessarily agree with your word choice, but you know I support a change in terms nonetheless. (My desire for different terms is due to the myriad definitions of PSTS and the resulting complications due to the law of primacy.) I think the wording will take a lot of time and effort to address, but it would be for the best in the end. After we get through this round of proposals, I'd be glad to work with you to try and work out a proposal for using different terminology.

I think narrowing the scope to purely "raw" sources is a significant change in and of itself. As such, it's another thing that should be addressed separately. I also believe it would remove many sources (from the "primary" category) that are a cause of concern in relation to NOR (such as historical and religious texts).

Point taken. I have added a footnote directing the reader to WP:V#Burden of evidence.

Thanks for the support and feedback. Are there other flaws in the draft that can be addressed? Is any of the wording problematic? Vassyana (talk) 03:26, 6 December 2007 (UTC)

Using sources

I propose adding the following to a section called "Using sources" immediately below the "Reliable sources" section:

Information in an article must be verifiable in the references cited. Article statements should not rely on unclear or inconsistent passages, nor on passing comments, even if the source is especially reliable. Passages open to interpretation should be precisely cited or avoided. Drawing conclusions not evident in the reference is original research regardless of the type of source. It is important that references are cited in context and on topic.

I don't think any of this is controversial and I believe it reflects what most people have voiced about sources and original research. Of course, I could be wrong, so let us know what you think. :o) Vassyana (talk) 01:04, 5 December 2007 (UTC)Text struck. See below. -- Text revised based on Fullstop's feedback.[3] -- Struck text removed.[4] -- Overt --> evident.

Alternate, with revised version of struck text (11:37, 5 December 2007 (UTC)). Italics are to note the additional text, nor for "live" use:

Information in an article must be verifiable in the references cited. Article statements should not rely on unclear or inconsistent passages, nor on passing comments, even if the source is especially reliable. Passages open to interpretation should be precisely cited or avoided. Drawing conclusions not evident in the reference is original research regardless of the type of source. When possible, cite only passages from the central topic of the source. A summary of extensive discussion should reflect the conclusions of the source's author(s). It is important that references are cited in context and on topic.

I've revised the text based upon Fullstop's feedback and looking towards a bit more simplification. I have also removed the struck text and provided an alternate reading with revised versions of the struck text. How are the revisions looking? Which version is preferable (if either)? Vassyana (talk) 11:44, 5 December 2007 (UTC)

Discussion

I think this is an accurate statement of current practice and good guidance for editors not familiar with choosing the best sources to support their claims. Something like this would be a positive addition to the policy. — Carl (CBM · talk) 02:59, 5 December 2007 (UTC)

Carl, would you still support the shortened version? Vassyana (talk) 03:07, 5 December 2007 (UTC)

Shorter policies are more likely to be read. The proposed passage is a mixture of NOR policy and advise about what portion of a source is most reliable, which probably belongs in WP:RS. The advice that does not pertain to NOR makes the article longer. It also makes the definition of NOR fuzzier (as if it isn't bad enough already); it tends to make NOR a synonym for sloppy writing. --Gerry Ashton (talk) 03:01, 5 December 2007 (UTC)

I've crossed-out the text that could be seen as more general advice. The remainder directly relates to original research. Vassyana (talk) 03:07, 5 December 2007 (UTC)

Without the that sentence, I think this is a valuable addition and consistent with current understanding and practice. ≈ jossi ≈ (talk) 03:20, 5 December 2007 (UTC)

In the end, the OR policy is intertwined with the verifiability and reliable sources policies, so some similarity is unavoidable. The shorter version is still accurate. The crossed-out versions are helpful, I think. They explain some of the criteria that are used in practice, in real articles, to decide which claims can be attributed to which sources. — Carl (CBM · talk) 03:23, 5 December 2007 (UTC)

(edit conflict)

Yep. Good job Vassyana. This is precisely the relationship of OR to sources. Nothing more. Nothing less.

For stylistic reasons (see following notes) I would write it as follows (sentence for sentence):

Statements in a Wikipedia article must be verifiable from the sources cited.

Avoid transmitting a passing comment or ambiguous sentence (open to interpretation) even when the source is a reliable source.

<strike third sentence since folded into second>

Drawing own conclusions or extrapolating a position is original research regardless of the type of source.

Where possible, use only those statements in the source that deal directly with the subject (being written for?).

A summary of an extended discussion must reflect the conclusions of the source's author(s).

Sources must be cited in context and on topic.
(alternatively: A source statement must have the same context and topic as the Wikipedia statement that uses that source.)

Notes:

I've switched from passive to active voice (eg "Claims left open" -> "Claims open") and avoided nested conditionals, etc.

I've avoided the word "claims" in favor of explicit mention of whether we mean source or target. This is ambiguous in the original, Eg. In sentence 3,4,5 its source while in 1,2 its target. Also "claims" sounds a little pejorative to me.

The word "transmitting" in sentence two should be replaced. I couldn't think of a better word.

The "even" clause that appears in sentence #2 is probably superfluous (ideally, all used sources are reliable). I've retained it anyway.

I'm not sure what "central topic of the work" in sentence #5 refers to (source or target), thus my "(being written for?)" in parenthesis.

Comments? -- Fullstop (talk) 03:25, 5 December 2007 (UTC)

I've matched struck out sentences now. -- Fullstop (talk) 03:28, 5 December 2007 (UTC)

Thanks for the feedback. I've revised the language somewhat in light of your comments. What do you think? Vassyana (talk) 11:45, 5 December 2007 (UTC)

I disliked the phrasing of "overt conclusions" in the first, but most of my concerns are well addressed by the rephrasing in the second version, and I support it. Phil Sandifer (talk) 14:31, 5 December 2007 (UTC)

I don't quite follow - both include the word 'overt'. The motivation for it is that the wording becomes too strict if it is replaced by 'stated' and too weak if it is replaced by 'obvious' or 'implicit'. If you can think of another word that fits better, that would be fine. — Carl (CBM · talk) 14:43, 5 December 2007 (UTC)

Ack, sorry - I missed that the "overt conclusions" sentence had survived. I would like to lose this sentence. I think it closes off too much. To my mind, there are three (very poorly defined) types of conclusion we can have in a source - overt, implicit, and, let's say, subtle/tenuous/extended/whatever. Overt, obviously, are OK. The latter category seem to me what we want to avoid. But I think implicit ones are also important, and, while not always acceptable, certainly not always wrong on the face of it either. (I'm using "implicit" here to refer, essentially, to obvious conclusions that the reader is meant to draw.) I don't know how better to phrase the sentence, and I'm inclined to simply remove it- I think the rest of the paragraph stands well on its own, and that the sentence is somewhat disposable within it. Phil Sandifer (talk) 14:49, 5 December 2007 (UTC)

Hmm, what about "evident" in place of "overt"? I think it's neutral enough to cover the acceptable spectrum of use, but still conveys the right idea. Vassyana (talk) 17:07, 5 December 2007 (UTC)

Evident works great. Phil Sandifer (talk) 17:13, 5 December 2007 (UTC)

Change made. Anything else that can be tweaked or improved that you see off-hand? Vassyana (talk) 17:22, 5 December 2007 (UTC)

Shouldn't incongruent be incongruous? Or if you really do mean incongruent (is there a fine distinction in meaning?), perhaps choose another word, as it's not as clear as it might be.  —SMALLJIM  14:33, 5 December 2007 (UTC)

I suppose "variant" or "discrepant" could be used instead of "incongruent". Any reasonable synonym indicating passages that are self-contradicting and/or dissonant with the overall tone/claims of a reference would suffice. Vassyana (talk) 17:07, 5 December 2007 (UTC)

After a bit of research I thought you were using this meaning of "incongruent", though I'm not sure now. Anyway based on the assumption (hope?) that I'm no dumber than the majority of WP editors, a simpler way of phrasing that concept should be used so that everyone can understand. Sorry, but "variant" or "discrepant" are not clearer, IMHO. It's important that all people who will be pointed to the page can easily understand it.  —SMALLJIM  17:49, 5 December 2007 (UTC)

Hmmm, "inconsistent"? I think it's a simpler word and still conveys the general idea. What do you think? Vassyana (talk) 18:23, 5 December 2007 (UTC)

Solution: replace "passing" with "incidental", then drop the comma before "or" and switch word order around to place "incongruous" last.

The sentence then reads:

Article statements should not rely on unclear, incidental or incongruous comments, even if the source is especially reliable.

Reason:

An incongruous/incongruent comment is a comment that is at odds with something. The unspecified "something" is the problem here, but is really only obvious because "incongruent" is immediately followed by a <comma> <adverb/adjective> <subject>

In the form it appears now, the clause with incongruous/incongruent needs a "with."

This can be made evident by replacing "incongruent" with a synonym: e.g. "... incompatible with ...," "... in disharmony with ...," "... in disagreement with ...," "... at variance with ...," "... at odds with ..." etc.

The rest is fine.

-- Fullstop (talk) 19:04, 5 December 2007 (UTC)

OK. Not sure now about the apparent emphasis given to "comments". So what about something like:

"==Using sources==

Information in an article must be verifiable in the references cited. Statements made in articles should not rely on unclear or inconsistent passages, or on incidental comments in those references, even if the source is especially reliable. …"

or even, reorganising a little:

"==Using sources==

The references cited in an article must verify the statements made in that article. However, these statements must not rely on unclear or inconsistent passages in the references or on incidental comments in them, even if the source is especially reliable. …"

(but maybe that's changing too much)  —SMALLJIM  20:05, 5 December 2007 (UTC)

This dispute is disputed.

I just had to comment on this. This has to be the most asinine possible thing for the page to get protected over.--Father Goose (talk) 06:49, 5 December 2007 (UTC)

The best thing to do with 'this is disputed' tags is to ignore them and work to resolve the actual dispute. Adding the tag doesn't change the content of the page, and removing it doesn't actually resolve disagreements. — Carl (CBM · talk) 13:20, 5 December 2007 (UTC)

I'll admit, I thought the same way as Father Goose when I first encountered this discussion. What changed my opinion (besides the fact all of the parties were making thoughtful contributions) is the chief challenge to any version of this policy, which is that it must be understood by people who have never written a research paper, & thus they have no idea why the rules for writing a Wikipedia article are different. <not entirely joking>I blame the educators: they claim that Wikipedia is full of unreliable information, but somehow overlook the fact it is created by the same people they educate.</not entirely joking> -- llywrch (talk) 17:33, 5 December 2007 (UTC)

First: don't blame the educators. :p And no, its not true that they claim that WP is full of unreliable information. What educators say is "do your homework!," which in the main means "read the damn sources yourself," and secondarily means "do not cite WP or any other encyclopedia." Encyclopedias are indispensable for background information.

Second: if students did not learn do the legwork themselves, how on earth can they be expected to write for the 'pedia? (cf. "it must be understood by people who have never written a research paper")

-- Fullstop (talk)

I think the truly asinine thing is that there is a heated dispute about whether or not the section is truly disputed. I guess people just don't see the irony. This is now the second time the article has been protected over the tag issue. COGDEN 22:42, 5 December 2007 (UTC)

Cogden, this whole dispute is a non-issue. No one is saying don't use primary sources. All people are saying is use them with particular care by sticking very closely to what the sources say, because they're easy to misuse. Why would you object to this? SlimVirgin ^{(talk)(contribs)} 00:04, 6 December 2007 (UTC)

Slim, that's just not true. The PSTS section used to say not to use primary sources. That is what brought the slow simmer over PSTS to a boil. Dhaluza (talk) 02:01, 6 December 2007 (UTC)

Can you provide a diff for that Dhaluza? Before this dispute started, the PSTS section stated:

Primary sources that have been published by a reliable source may be used in Wikipedia, but only with care, because it is easy to misuse them".

This was the status quo for a very long time. The real dispute has been over further loosening and even removing the limitations on use of Primary sources. Dreadstar † 02:20, 6 December 2007 (UTC)

Plus, in some cases, primary sources are still entirely banned. Highly-technical primary sources are banned, because nobody without specialized education can verify to them, while, anomalously, highly-technical secondary sources are allowed, even though nobody can understand them other than Ph.Ds, and even though the secondary sources are, themselves, also primary sources. COGDEN 02:20, 6 December 2007 (UTC)

As I argued several sections ago (in the section about the hypothetical highly technical math article drawn from a primary source), I don't think it's necessary to interpret the policy this way. But, frankly, it's a reasonable approach for a general encyclopedia such as Wikipedia ("the encyclopedia that anyone can edit"). If the material is called into question and WP's most highly technical editors can't find a reliable secondary or tertiary source that says the same thing they assert is in the primary source, then don't use the primary source of the highly technical material and remove the material in keeping with the policy. ... Kenosis (talk) 03:26, 6 December 2007 (UTC)

I also think the protection was over the top. As far as I could see, everyone was keeping to the one-revert-rule, so that is hardly an edit war. By protecting the page, we never see how many people are willing to go on the record for one version or another. And it blocks further work on other issues. Dhaluza (talk) 02:01, 6 December 2007 (UTC)

You should refresh your memory about edit warring. If people want to be on the record, they can make a statement on this talk page. — Carl (CBM · talk) 02:11, 6 December 2007 (UTC)

We have discussed this on the talk page, and the editing was in conjunction with talk page discussion. So cutting off this process is hardly conductive to reaching consensus. And how is protection better? If the only "problem" is adding and removing the tag, what is the harm? Since there is no agreement over whether it should be in or out, how does arbitrarily selecting one alternative solve the problem? I would submit that having it on the page part time is less disruptive than locking it in or out. Dhaluza (talk) 02:19, 6 December 2007 (UTC)

It was an edit war, plain and simple; so the protection was right on target and needs to stay in place until this entire dispute is over and done with. I do not believe that Policy pages should contain dispute tags. Dreadstar † 02:25, 6 December 2007 (UTC)

Tags are not pretty, but they are far better than having perpetual protection and a policy article that claims to have widespread consensus, but really doesn't, while a discussion on the talk page rages on for eternity, consensus gradually falling further and further away from the protected language, and there's nothing we can do about it until the original cabal of authors dies. COGDEN 02:40, 6 December 2007 (UTC)

I agree with you, right up until the end. Please try to be patient, and civil. Dhaluza (talk) 10:42, 6 December 2007 (UTC)

The primary source section has been in the policy for three years

To people who are arguing that this is a new(ish) section, or that the meaning was recently changed, the stress on secondary sources was added on December 10, 2004 (see lead), [5] and entered the policy in more or less its current form on March 8, 2005, when it said:

"In order to avoid doing original research, and in order to help improve the quality of Wikipedia articles, it is essential that any primary source material used in an article has been published or otherwise made available to people who do not rely on Wikipedia. Moreover, it is essential that any generalization, analysis, synthesis, interpretation, or evaluation of information or data come from a secondary source that is available to readers (e.g. in a library or non-Wikipedia web-page). [6]

Three years means that this is very much the consensus version, and it really can't be changed or kept tagged because one or two people don't like it. All policies would be in a constant state of war were that the case.

Cogden, it strikes me that you must have had a particular experience of trying to add primary-source material somewhere, and being prevented by this section, for you to be as keen as this to change it. Can you show us what that issue was, please? It's possible that the policy was simply misused in your case, and that this is all a giant misunderstanding. Wouldn't that be nice? :-) SlimVirgin ^{(talk)(contribs)} 07:24, 6 December 2007 (UTC)

Amen to that, Reverend Mother. And, in anticipation of COGDEN's reply, let's not forget that when an ad hoc policy change closes a door, somewhere it opens a window. Avb 11:17, 6 December 2007 (UTC)

SlimVirgin, you have made this "really long time" assertion before, and I have refuted it with the diffs, so here they are again:

This edit from March 8, 2005 introduced the original definition of primary sources as factual sources. In this edit from October 23, 2006 you significantly changed the definition to close sources.

Frankly, I don't think there is any significant disagreement with the quote you cited above. Nobody has made a serious argument for allowing unpublished primary sources, and there has not been any serious argument about excluding editors' analysis or synthesis over those published by RS. So if you are proposing to trim PSTS down to these essential points, I would support that. The only thing I question is the need to use the defined terms primary and secondary sources. Just leave that out and you get this even more succinct and direct version:

"In order to avoid incorporating original research in Wikipedia articles, source material must have been previously published or otherwise made available to people other than on Wikipedia. Moreover, it is essential that any generalization, analysis, synthesis, interpretation, or evaluation of information or data come from a source that is available to readers (e.g. in a library or non-Wikipedia web-page)."

-- Dhaluza (talk) 11:02, 6 December 2007 (UTC)