Wikipedia talk:No original research

RfC: Policy on primary sources is too extreme

WP allows primary sources to support isolated facts. Comment is requested here upon extending WP:V and WP:Primary to explicitly allow entire articles based upon primary sources, provided these articles contain no synthesis, analysis, or interpretation of the original sources. A summary of a recent discussion follows this RfC at Wikipedia talk:No original research. Brews ohare (talk) 18:46, 8 February 2012 (UTC)

Summary of discussion on this Talk page

I'll try to summarize the views expressed.

Topic The issue was raised that statements from WP:Primary,

Material based purely on primary sources should be avoided.

According to Wikipedia:Notability

if no reliable third-party sources can be found on a topic, then it should not have a separate article

and also, according to WP:V:

Base articles largely on reliable secondary sources. While primary sources are appropriate in some cases, relying on them can be problematic.

These statements are too strong because they do not allow any articles that are made up entirely of primary sources. In WP:V there are these conditions upon using primary self-published sources:

1. the material is not unduly self-serving;
2. it does not involve claims about third parties;
3. it does not involve claims about events not directly related to the source;
4. there is no reasonable doubt as to its authenticity;
5. the article is not based primarily on such sources.

The opinion expressed was that the first four of these requirements for self-published sources sufficed to prevent misuse of articles containing only primary sources, and the last restriction was not needed, and indeed would reject some useful articles.

Educational comment: "Primary" is not another way to spell "self-published". This entire proposal is based on ignorance of what a primary source is. See, e.g., the extensive quotation of WP:SPS, followed by assertions that this is what WP:V says about primary sources. I have struck the erroneous assertion that WP:V imposes these conditions on primary sources and added the accurate bold-faced words to the previous statement. WhatamIdoing (talk) 18:41, 9 February 2012 (UTC)

Responses Response to these observations were negative, but not directly relevant to the issue. They fall into these categories:

No useful articles like this exist, even in principle, so it is a non-issue. There are such articles, and an example is given later.

If an article like this comes up, it can be dealt with as a possible exception to policy, and no policy change is needed. This is a utopian position; exceptions to policy are very rare. And there is no need to force this situation: just change the policy.

If an article like this comes up, secondary sources always can be found to replace or supplement the primary sources, so the question comes down to whether the article is "primarily" based upon the primary sources. It is not true that secondary sources always can be found; books are written with different objectives than WP articles.

I find none of these objections addresses the point raised, namely there is no need for the blanket refusal to accept articles based entirely upon primary sources, and the four first items in the list of reservations above suffice without the fifth reservation. To assist in this consideration the following example was provided:

Example straw man

an article with the subject "State requirements for the office of district attorney" with subsections: "Arizona requirements", "California requirements", and so on for all the states.

If the present policy were strictly enforced, a secondary source that describes each state would have to be found, and only then could a primary source be used to supplement the secondary source(s). The policy has the nonsensical effect of requiring a secondary source and citing: "The book Law in America says 'Arizona law says x (Ref: Arizona law)' " instead of saying directly 'Arizona law says x (Ref: Arizona law)'. WP would not require that for this specific fact, so why should it be required for an article consisting only of such specific facts, with no interpretation, analysis, or synthesis?

In fact, there is an article already on WP United States Attorney, that is somewhat like this, but does have some secondary sources, and present policy raises the non-issue of whether it is primarily based upon primary sources. Such a consideration should arise only under the circumstances described in the first four restrictions cited above.

Are there further comments on the proposal to drop the last-named restriction upon articles consisting only of primary sources?

Comments

If you are not satisfied with the comments above in this same talk page you can raise an WP:RfC about it. Dmcq (talk) 17:54, 8 February 2012 (UTC)

Hi Dmcq: That sounds like a good idea. I have put in an RfC template

• Comment: This change in policy will avoid tendentious articles by excluding synthesis, interpretation, and so forth, to allow only strict reporting of the content in primary sources. This policy change will allow articles in WP that are simply descriptions of what sources say about themselves. Examples are articles like WP:Formal organization or United States Attorney. A policy change avoids arguments about whether such articles are "primarily" based upon primary sources. It also avoids the need to find a secondary source S to say things like "Secondary source S says 'Jimmy Wales is on the board of the Wikimedia Foundation' (See primary source Wikimedia)" and instead to say directly 'Jimmy Wales is on the board of the Wikimedia Foundation' (See primary source Wikimedia). Brews ohare (talk) 18:57, 8 February 2012 (UTC)

• Comment: The aspect of notability is missing from all this, and without it, it opens the door to huge number of articles of non-synthetic, straight-up summary of primary sources but without any context. The impression I get from this discussion is we are mostly talking about list-style articles that summarize in a non-synthetic manner several disparate primary sources. It is not that it is bad to do this, but the problem is that there needs to be some reason why we have this list. In the specific example, one certainly can find enough secondary sources to explain what a DA is, and that the route to becoming a DA are diverse by state; once it is established that the routes and means to becoming a DA is a notable topic, primary-sourcing the individual state policies makes perfect sense. But if the list as a whole could not show notability through secondary sources, then the list shouldn't exist in a stand-alone form.
  There's a lot of other complex interactions here when you move away from list-style articles, but still, we do expect every article to show notability, and thus show some amount of secondary sources. The ratio of primary to secondary may be vastly skewed depending on the topic (again, for the DA example, with 50 states, that's 50 primary sources, while I'd only expect to have to see 3 or 4 to establish the notability of the pathway to becoming a DA - that is not a problem in terms of this discussion). --MASEM (t) 19:10, 8 February 2012 (UTC)

Masem: I'd agree that some such articles may not be notable, but if they should not be on WP because of that, doesn't Wikipedia:Notability get rid of them? Shouldn't we reject articles with other faults for their actual failings, not because they refer to primary sources? Brews ohare (talk) 19:44, 8 February 2012 (UTC)

Notability requires coverage beyond primary sources. Mind you, I recognize the complex issue where we do know there's secondary coverage but prefer the primary as it is less interpretive (such as the process to be a DA by state), so it is unlikely that notability would affect this type of article. But when only primary sources are available - no secondary ones at all - then it is failing notability so deletion would be appropriate. --MASEM (t) 19:57, 8 February 2012 (UTC)

Notability as written does require deletion, but it also should be rewritten. The important issue here is WP:Indiscriminate, which covers WP is not just a collection of information. Again, shouldn't we reject articles with other faults for their actual failings, not because they refer to primary sources? Brews ohare (talk) 20:06, 8 February 2012 (UTC)

The actual failing of an article that only can be sourced to primary source is that there's no evidence of why it is important to include in a work that is not an indiscriminate collection of information. For example, from the video game area, I could easily make an article about the weapons in a video game, all sourced from the game and strategy guide. But unless sources go into any secondary, analytic depth about the weapons, this information has no importance to the larger picture of being an encyclopedia. --MASEM (t) 20:29, 8 February 2012 (UTC)

Your example can be eliminated from WP using WP:Indiscriminate. An article like WP:Formal organization uses only primary sources but it is useful as a guide to the many pages describing WP organization. An article like the "straw man" above could be similarly useful. Of course, this information could be used for other purposes, in articles drawing comparisons, evaluations and so forth, and those would require secondary sources. Why be restricted to articles making judgements and assessments?

Why assume absolutely no useful articles exist that are based upon primary sources alone? Why not let the applicable policies like WP:Indiscriminate be applied to eliminate inappropriate articles? Why assume we are smart enough to anticipate all possibilities, and rule this class of articles out altogether? Brews ohare (talk) 20:48, 8 February 2012 (UTC)

We need to establish why the topic is relevant to the reader. Without this, it is just datum, and not appropriate to the encyclopedia, as well as a conditioned that can be gamed. To show a topic to be relevant , we need secondary sources to show that's the case, otherwise, it is just original research.

I would argue that most articles that you point to that are presently only sourced by primary sources can be augmented with a few choice secondary sources in the lead to establish that point, ergo establishing notability and of course moving the article from being only sourced to primary sources, avoiding the issues of deletion/merging. --MASEM (t) 21:19, 8 February 2012 (UTC)

Please see my remarks to Dmcq below. Brews ohare (talk) 22:45, 8 February 2012 (UTC)

An article like WP:Formal organization uses only primary sources...

No. There is no such article. There is a project page, but no article. See the "WP:" at the start? That means "not an article". WhatamIdoing (talk) 18:44, 9 February 2012 (UTC)

• Comment: Well knowing which books are on which floor in a university would help Wikipedia's educational aims, but should I really be putting in an article about every single library saying what the layout of its bookshelves is? And there are lots of recipe books around, surely being able to find a recipe easily is good and these are in books so shouldn't I stick their contents lists into Wikipedia? And how about the current train timetables? That'll get students to class in time. We really do need secondary sources to stop Wikipedia being filled with stuff that just isn't notable no matter how useful it is. A secondary source is visible evidence somebody finds the stuff notable rather than editors doing original research and deciding things like that for themselves. There is no weight without a secondary source. Dmcq (talk) 20:00, 8 February 2012 (UTC)

Dmcq: You have identified some possible articles that use only primary sources, and should not be on WP. However, WP:Indiscriminate is the relevant policy here: WP is not just a collection of information. I'd suggest that there are some articles of this type that are in fact useful, and rather than be clairvoyant and dismiss all such articles blindfolded, we should dismiss them for their actual failings, using pertinent policies, not because they are based only upon primary sources. Brews ohare (talk) 20:12, 8 February 2012 (UTC)

That is exactly my point. How do you show something is indiscriminate? I gave explanations of why each of my examples was useful and furthered the educational purpose of Wikipedia. Who are you to say that I'm wrong? We could have AfD's all day and night with editors fighting over tons of things like this without some decent bright line. And that minimal requirement is notability as shown by secondary sources otherwise it is a fairly quick exit. Dmcq (talk) 21:15, 8 February 2012 (UTC)

Dmcq & Masem: The purpose for such an article could be (for example) to act as a map locating the topics or sources outlined. Additionally, the article should contain enough summary detail to enable a reader to find what they are looking for. That is what WP:Formal organization does.

The statement of purpose doesn't have to be from a secondary source. In fact, a secondary source for (say) a guide to some classification of web sites, might be hard to find, even if the guide were useful.

If it is difficult to decide whether the statement of purpose is sufficient for inclusion, then maybe that debate should occur. Of course, there may be no unanimity over the value of "Regulations regarding dogs and rabies in American cities". The debate would decide if there was really any interest in the topic. Acceptance shouldn't be decided simply because all the sources were to government web sites.

Possibly, the proposed revision of policies should include a requirement for a "statement of purpose" and provide some guidelines for an acceptable statement? Over time a sensible set of guidelines would evolve, don't you think? Brews ohare (talk) 22:45, 8 February 2012 (UTC)

If we are talking specifically mainspace articles that would fall as "outlines" or the like, we generally do not apply notability guidelines to them, in the same manner we don't apply them to disambiguation pages - any page that helps in general navigation is ok. As part of the larger picture on that, however, if you are creating a navigation page for a very broad topic, it is absolutely certain that the topic is notable.

A statement of purpose does need to be secondary, otherwise you enter into the realm of COI to maintain an article on WP.

But let be me clear: there is a difference between an article that currently has only primary sources but it is known that secondary sources exist that repeat the primary sources; and an article where only primary sources exists and no secondary sources have been found outside of the article itself. The former we'd generally keep, because as long as there's some secondary interest, that's evidence of notability.

Your example of dogs and rabies raises a different problem. The subject of dog and rabies regulations is notable, and you probably can source the local regulations for them and fill the article with primary sources. But would we ever allow such an article? Likely not, not due to sourcing but due to it being indiscriminate - discussing the specific regulations per town (10,000s of them) is excessive for a encyclopedia that is meant to summarize. An article discussing the types of regulations in broad strokes, borrowing examples from specific regulations using primary sources, would be more appropriate here. But this is not an issue because of the primary sources, just the shear number of entries. --MASEM (t) 23:05, 8 February 2012 (UTC)

• Comment: Masem: You suggest that mainspace articles that are outlines are not usually subjected to WP:Notability, which is good news. However, this policy does state:

"If no reliable third-party sources can be found on a topic, then it should not have a separate article"

So the fact WP:Notability is seldom used this way probably is because WP:V and WP:Primary come to editors' minds more readily. I'd take it that you would be agreeable to explicitly exempting 'outlines or the like' from these strictures against articles relying upon primary sources alone? Is that so? Brews ohare (talk) 14:05, 9 February 2012 (UTC)

My read of the various guidelines that talk about outlines and other formal organization mainspace "articles" used mostly for navigation is that they are generally not subject to the same content and inclusion policies as they work on the assumption that the outline is not introducing anything new to WP; however, there is disagreement if WP should support these types of "articles" that goes beyond WP:OR and other content policies, as well as WP:N and inclusion guidelines. For purposes of this discussion I would exclude discussion of outline and navigation-style articles since how policy and guidelines apply to them are different, and instead focus on articles with new content and how primary sources apply to them. --MASEM (t) 14:27, 9 February 2012 (UTC)

Masem: I'd appreciate some guidance to find these guidelines exempting outlines and such from WP:Primary, WP:Notability, and WP:V. It may be that these exemptions are really all I am looking for here, and some language in WP:Primary, WP:Notability, and WP:V referring to these guidelines would be adequate to my concept of what is needed here. Brews ohare (talk) 14:38, 9 February 2012 (UTC)

Well, the problem I see with that is that the example being used, the article on the various state requirements for becoming a DA, would not be an outline or similar type of article and that seems to be the type of article we're focusing on. So I don't think this advice is going to help with the concept you're trying to set. --MASEM (t) 14:57, 9 February 2012 (UTC)

Maybe the language you are aware of could be helpful? It seems that you think the definition of "outline or the like" needs extension somehow. I'd guess we could come up with some kind of trial balloon? Brews ohare (talk) 15:10, 9 February 2012 (UTC)

I have begun such a trial balloon approach in a comment further down the page. Please take a look. Brews ohare (talk) 15:50, 9 February 2012 (UTC)

• Comment:

  [Primary sources] are very close to an event, often accounts written by people who are directly involved, offering an insider's view of an event, a period of history, a work of art, a political decision, and so on... a scientific paper documenting a new experiment is a primary source on the outcome of that experiment. Historical documents such as diaries are primary sources.... Any interpretation of primary source material requires a reliable secondary source for that interpretation.

It seems to me this rule says it all. If, for example, Wikipedia wanted an article about Pepys' diary and used only the diary from its initiation, in simple terms that would merely be one of two things: a pure summarisation using only the diary, or, direct quotes from the diary in a piecemeal way. Either usage is OR without secondary sources, and that is why I think article creation rules should remain as they are now. May I offer a real counterexample with Birthstones: several lists are in the article, including the Gem trade and Jewelers (both of America). However, without secondary and tertiary sources behind them, the article would only look like a commercial advert for the two associations, nothing more. As it is, the article was created as a well-rounded and encyclopedic article about birthstones. A good specific example I can cite is mineralogist Georg F. Kunz's work, which is perfect as secondary reference on the subject and is cited in the article. It cannot stand without Kunz or its other secondary citations- but if one were going to make an article about birthstones using only one source, that seems to me equivalent to the change being proposed here. For all those reasons I oppose this suggested change to the rules.--Djathinkimacowboy 05:23, 9 February 2012 (UTC)

Djathink: You raise an important point, a possible misuse of primary sources to provide a parochial view of a subject by reporting only part of a subject through primary sources. I'd respond to this objection in several ways. First, all articles evolve, and if a first presentation of a topic is based too narrowly on primary sources, those recognizing this fact can add secondary sources (if there are some) or flesh out the presentation with other primary sources, or more balanced quotations from the primary sources themselves. Second, if objections to the article can be raised using criteria of WP policies that are not related to primary sources, those policies should be applied, not the objection to use of primary sources. For example, WP:NPOV comes to mind. Third, there are in place already important restraints upon the use of primary sources quite apart from the condition that not only primary sources be used. I think these conditions (which roughly speaking disallow interpretation, synthesis, and evaluation) should suffice to control major abuse of primary sources. The aim here is not to allow unbridled use of primary sources, but to remove the restriction that "no article shall rely primarily upon primary sources." It may be that removal of this restriction will require additional stipulations upon use of primary sources: that seems fine with me. I simply feel that the present "no article shall rely primarily upon primary sources" is a blunt instrument where a scalpel should be used instead. Brews ohare (talk) 13:54, 9 February 2012 (UTC)

• Comment: Summaries of wide reaching government legislation can support an article on there own. Mundane, non-controversial but important legislation may not spawn secondary sources for a time, but a summary of such legislation is important encyclopedic knowledge. When discussing the "asserted position/policy" of a person or group, a primary source is the most reliable source and should be used over secondary sources. Content without notability, etc, is best handled under its respective policy. wp:primary should not be used as a tool to enforce policies covered elsewhere.Gsonnenf (talk) 12:48, 9 February 2012 (UTC)

Are you really saying governments can pass wide reaching government legislation of note without anybody secondary sources commenting on it? Dmcq (talk) 13:20, 9 February 2012 (UTC)

The key point here is the last statement, which is the thrust of this entire thread: WP:Primary should not be used as a tool to enforce policies covered elsewhere. Or perhaps, WP:Primary should not be used as a substitute for enforcing other policies. Brews ohare (talk) 13:34, 9 February 2012 (UTC)

If you want to talk about removing the definition of primary out of the Original research policy that's something else but currently this policy is the place to go for finding what primary sources are. That part is integral to how primary sources can be used. It would help nobody and it certainly wouldn't support editors if we have to go round the houses trying to build that up from the various policies instead of stating it in a straightforward statement in a description of primary sources. The asserted policies of an organisation are only of interest if its policies are of interest and you need secondary sources for that. I think people may be mixing up verifiability and citation here, citations are only needed if verifiability is disputed. Dmcq (talk) 14:10, 9 February 2012 (UTC)

Dmcq: I don't think this thread is about changing the definition of primary sources. It is about replacement of the present language that says more or less "no article is acceptable that uses only primary sources" with language that allows such articles subject to the other restrictions already in place. These strictures include those against interpretation, evaluation, or synthesis. Maybe if "no article is acceptable that uses only primary sources" is deleted, some different wording will be needed. Any suggestions? Brews ohare (talk) 14:17, 9 February 2012 (UTC)

For example, one suggestion arising from Masem's comments would be to replace "no article" with a list of articles that could be used this way: "outlines and the like are acceptable that use only primary sources" subject, of course, to other WP policies and to the requirements for no interpretation, synthesis, or evaluation of the primary sources. Brews ohare (talk) 14:27, 9 February 2012 (UTC)

Dmcq, if you feel something is not notable you can bring it up on a talk page and delete it under WP:NOTE. Some topics with secondary sources are not notable, some topics based on primary sources are notable. Editors should not be forced to hope for some obscure secondary source, such as a brief confirming article that a mundane non-controversial policy was passed, when there is consensus that the primary source is far more authoritative and notable then the secondary source. Having a mix of poorly focused rules enforcing a distinct goal (notability) will only encourage people to game the system and create bizarre intertwined wikilawyer arguments. Gsonnenf (talk) 02:18, 10 February 2012 (UTC)

• Comment: Here are some possible approaches to relaxing the present restriction that "no article is acceptable that is based upon only primary sources":

• Approach (1): Subject always to all WP policies regarding synthesis, point-of-view, interpretation, evaluation, notability and so forth, articles using only primary sources are allowed, but of course may be amended at any time to include secondary sources or additional primary sources. It is anticipated that most articles of this nature are preliminary, and will become full-fledged articles in time.
• Approach (2): Subject always to all WP policies regarding synthesis, point-of-view, interpretation, evaluation, notability and so forth, articles using only primary sources are allowed that are intended only to facilitate the navigation of sources pertaining to a topic. As such the sources may be quoted or outlined strictly for the purpose of guidance to the content of the sources, to assist a reader to find their way to the sources that are pertinent to their interest.

Approach (1) is very open-ended and simply amounts to dropping the restriction altogether, subject of course to other policies.

Approach (2) is restricted, and replaces the outright ban with a more specific allowance of certain types of article. The specification of allowed articles could be made more restrictive or less restrictive: maybe some suggestions could be entertained? Brews ohare (talk) 15:44, 9 February 2012 (UTC)

Brews... Your proposals contain an internal contradiction. You say: "Subject to <WP:Notability> articles using only primary sources are allowed". The problem is WP:Notability says: "Sources, for notability purposes, should be secondary sources, as those provide the most objective evidence of notability. In other words WP:Notability requires at least one secondary source, and therefore indicates that articles using only primary sources would NOT be allowed. To restate your proposals, you are saying: "Subject to WP:Notability (which requires at least one secondary source) it is ok to write an article with no secondary sources". Contradiction! Blueboar (talk) 16:14, 9 February 2012 (UTC)

That's a good way of putting it. If an article is of the type subject to notability requirements, there is no way it can get by on primary sources alone within the article. If the article is of the type that notability does not apply (which presently is only includes disamb pages), then primary sources alone are fine, but usually these types of articles don't need to be sourced anyway since they're navigational in purpose. --MASEM (t) 16:24, 9 February 2012 (UTC)

Blueboar and Masem: Excuse any confusion on this point. As outlined at the outset of this RfC, it is proposed to change the wording regarding this matter in WP:V, WP:Primary, and WP:Notability (and any other places where this restriction pops up). As you point out, it makes no sense to change one policy and not the rest.

Masem: I gather that you believe WP:Notability does not apply to articles of the type described in Approach (2). However, WP:Notability's categorical refusal:

"If no reliable third-party sources can be found on a topic, then it should not have a separate article

does not explicitly exempt navigational articles. Perhaps changes in all the governing policies would suffice that specifically exempt navigational articles from a requirement for secondary sources? Brews ohare (talk) 16:32, 9 February 2012 (UTC)

Again, Secondary does not mean third-party. You need to find out what the term primary actually means before you embarrass yourself any further. Eliminating a requirement for secondary sources will not have the smallest effect on the separate and unrelated requirement that WP:Independent sources have taken notice of the subject. WhatamIdoing (talk) 18:49, 9 February 2012 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ WhatamIdoing: Of course, you are right that these two things are different. To restate matters more clearly:

Perhaps WP policies can be rewritten to allow navigational articles to be based entirely upon primary sources. This rule is an example of what I have called Approach (2) above.

Brews ohare (talk) 19:02, 9 February 2012 (UTC)

It may be best if you list out some examples what you think are "navigational articles" where this advice on primary sources would be intended, as I have a feeling there's a disconnect here. That may help to clarify what changes you want. --MASEM (t) 19:36, 9 February 2012 (UTC)

Masem: The trial balloon for a "definition" is:

A navigational article is one intended only to facilitate the navigation of primary sources pertaining to a topic. In a navigational article, the primary sources may be quoted or outlined strictly for the purpose of guidance to the content of the sources, to assist a reader to find their way to the sources that are pertinent to their interest. As with all articles, any interpretation, evaluation, or synthesis of views in the primary sources must be substantiated by secondary or tertiary sources.

How's that? Brews ohare (talk) 19:48, 9 February 2012 (UTC)

First, I think it would be helpful to distinguish between an "article" and a "page". The purpose of an article is to present information on a topic. The purpose of a navigation page (such as a dab) is not to present information - it's purpose is to aid readers navigate Wikipedia and find articles on the topics they are looking for.

That said - There is no need to rewrite our content policies to allow navigation pages to be based entirely upon primary sources... because navigation pages are not articles, and our content policies only apply to Articles. There is no reason to allow navigation pages to be cited to primary sources, because navigation pages are not required to cite any sources... Navigation pages are not subject to our content policies. . Articles require sources, but navigation pages are not articles.... they are simply aids to locate articles. Blueboar (talk) 19:50, 9 February 2012 (UTC)

Blueboar: We are running into semantics here. The above is to be viewed as a definition of a completely new category of article. Perhaps we need a new designation. How about this:

A familiarization article is one intended only to facilitate the acquaintance with primary sources pertaining to a topic. In a familiarization article, the primary sources may be quoted or outlined strictly for the purpose of guidance to the content of the sources, to assist a reader to become familiar with the sources that are pertinent to their interest. As with all articles, any interpretation, evaluation, or synthesis of views in the primary sources must be substantiated by secondary or tertiary sources.

The follow up on this is the statement in relevant policy pages that familiarization articles can be based entirely upon primary sources, but if any interpretation, evaluation, or synthesis of views in the primary sources is presented, it must be backed up by secondary or tertiary sources. Brews ohare (talk) 20:02, 9 February 2012 (UTC)

Suggestion: Brews, distinguished body of everyone here, what if someone could distill the best of what has been posted herein and use that to alter the language of the rule slightly? That wouldn't exactly constitute a major change in policy, and it would have the happy benefit of doing what you mentioned about articles: deftly applying a scalpel instead of an ICBM. My small opinion is that the discussion is getting a bit off the rails....--Djathinkimacowboy 07:40, 10 February 2012 (UTC)

The only thing that has come up that I can see is that disambiguation pages aren't subject to the normal rules which is something everybody assumes anyway. There is no secondary source talking about the notable topic of disambiguating Waterford in Ireland from Waterford in Australia or Canada or England or the United States or all the other uses. There is no reason to start sticking something in the OR policy about primary and secondary sources because of this. Is there some point about this discussion about primary sources as far as a straightforward ordinary article is concerned? I don't see it. Dmcq (talk) 10:39, 10 February 2012 (UTC)

Comment I can see a case for an article to be based on a primary source, but backed up by secondary sources. Such an article that is on my "to-do" list is the armoured train ambush during the Boer War in which Winston Churchill was captured. The de facto reference works regarding the ambush are Churchill's own writings (he was in South Africa in 1899 as a journalist). The notablility of the ambush only arose once Churchill became Prime Minister (40 years later). Subsequent writings have been an analysis of what Churchill wrote rather than a rewrite of the event. Martinvl (talk) 16:32, 10 February 2012 (UTC)

We wouldn't be interested in that unless a secondary source had commented on them and an analysis by a secondary source can certainly show what are the important points in his writing. Everything is based on primary sources at the end, that does not mean that articles are solely based on primary sources because secondary sources have pointed them out and commented on them. Writing a section in the article about something in his book that nobody has shown any interest in though would be definitely something to avoid according to this policy. Dmcq (talk) 22:26, 10 February 2012 (UTC)

• Comment: Allowing to create (scientific) articles completely based on primary sources would destroy any scientific credibility of WP. We should rely on the established opinions as published in accepted textbooks or review articles in reputable journals, because this is the only way to establish WP:Notability and WP:Undue. Primary sources might probably be used in addition to secondary sources - but only the latter should be used as the foundation of an article --D.H (talk) 11:29, 12 February 2012 (UTC)

"familiarization article"

Brews, I have a problem with the entire concept of a familiarization article. First, what you seem to be envisioning is a stand-alone annotated index of sources. I don't think an encyclopedia is the right place for that.

But, let us assume that I am in the minority, and others think it is a great idea... you would still need a secondary source to establish notability. Every stand alone article has a specific topic that is unique to that article. In a "familiarization article" as you envision, that topic would be: "Primary sources pertaining to X". OK... So what makes the topic of "Primary sources pertaining to X" notable enough for its own stand-alone article? You would need to establish notability... and to do that, you would need a secondary source that discusses the concept of "primary sources pertaining to topic X", or at best discusses these sources as a distinct group. Blueboar (talk) 14:59, 10 February 2012 (UTC)

Notability requirements could be changed for these articles, or dropped. For example, what is the notability of WP:Formal organization or any information page on WP like WP:Mathematical symbols? Brews ohare (talk) 15:15, 10 February 2012 (UTC)

Content and inclusion guidelines do not apply to any Wikipedia-space articles. These are non-examples. --MASEM (t) 16:05, 10 February 2012 (UTC)

The only reason WP:Formal organization is not part of the Main space article Wikipedia is that many editors invoked WP:Primary to deny its presence there. Brews ohare (talk) 16:18, 10 February 2012 (UTC)

I'm going to say it; that specific page is highly exceptional, and impossible to categorize in the overall theme of primary vs secondary sources, because it involves navel gazing, COI issues, and the like. Basically, any page in mainspace about Wikipedia is going to be tricky to write about, and why we generally fall back to secondary sources to assure that the article can be written (eg Reliability of Wikipedia.) We can't use that article to establish any precedent here. --MASEM (t) 16:30, 10 February 2012 (UTC)

Hi Masem: I am glad to see that such a thing as an exception is out there. It raises the specter that there may be other exceptions, eh? Brews ohare (talk) 16:43, 10 February 2012 (UTC)

It may be that "navel-gazing" pages are not that uncommon. For example, one could use the description given by any other web site, like Citizendium Knowino in the same way. Also, the description of any corporation's or charity's hierarchy. Brews ohare (talk) 16:52, 10 February 2012 (UTC)

Wikipedia having an article about Wikipedia or any other Wikimedia Foundation sites can be taken as navel gazing without proper secondary sources. Wikipedia having an article on any other website like Citizendium or the like is not navel gazing - they are not part of Wikipedia/Wikimedia. We still require secondary sources to talk about that website in an encyclopedic manner. --MASEM (t) 17:30, 10 February 2012 (UTC)

If I understand your view, if one wished to describe the organizational structure of Citizendium and could find no source describing that structure other than CZ's own documentation, then no article could be placed on WP about this structure, and in fact, any outline of this structure as part of another article would have to be removed if challenged because it relied entirely upon a primary source?

That seems extreme. Brews ohare (talk) 17:42, 10 February 2012 (UTC)

The same position would disallow a description of the governmental structure of Mozambique (presently unsourced entirely) or Chad (presently sourced to the BBC, seemingly less authoritative than Chad itself), or the organization of the presidential primary caucus in Nevada. Brews ohare (talk) 17:48, 10 February 2012 (UTC)

If, using your example, there was no source describing CZ's structure outside of CZ itself, then we would not allow an article on it. That doesn't mean that that structure cannot be talked about in context of the article about CZ, as your other examples do. It is not the case we don't allow primary sources, it is that we simply don't allow articles solely based on primary sources. --MASEM (t) 18:00, 10 February 2012 (UTC)

Masem: Surely that is silly? So if I want to write about CZ organization I can draft the article: "Organization of on-line encyclopedias" with a subsection "Citizendium" that is exactly what would be contained in the stand-alone "Organization of Citizendium" and that is fine? And likewise, in an article on Chad I can discuss its governmental structure sourcing the BBC News seven years ago, but I couldn't put the same thing in "Chad (Governmental structure)" sourcing documents from the government of Chad? Nonsense, I'd say. And something that should be changed. — Preceding unsigned comment added by Brews ohare (talk • contribs)

If you wanted to write about the CZ organization - which is only defined by CZ itself, you would go to the Citizendium (which is notable) and write about it in there; that's the more natural location for that. Chad's government, on the other hand, is covered by secondary sources and can be written about in a separate article. --MASEM (t) 18:47, 10 February 2012 (UTC)

Masem: So WP:Formal organization is more naturally located on Wikipedia? My thoughts exactly. We may be getting caught up in the examples instead of the principle. The principle (not the practice on WP) is that material that is suitable as a subsection of a larger article should not be disqualified from being split off as a separate article. If its support is OK for the larger article, it is equally OK for the stand-alone. If "Nevada caucuses" can exist in "Caucus" it can exist in "Nevada caucuses". Brews ohare (talk) 19:03, 10 February 2012 (UTC)

No, because the level of detail and information in WP:Formal organization is not aimed at a general reader, but for an editor of Wikipedia - it is necessary navel gazing within Wikipedia namespace to allow editors to understand how the work is organized. For the average reader, the organization in general is discussed currently within the Wikiepedia article; no further article is needed. And yes, actually, while material can be included in a sub-section of an article, splitting it off is not appropriate per WP:N and WP:SPINOUT. Only notable topics should be spun out to separate articles. --MASEM (t) 19:15, 10 February 2012 (UTC)

Masem: You are framing the situation differently. Splitting off may be inadvisable for various reasons. I agree that the intended reader of an overall article may have limited interest in a lot of detail that might be appropriate in a stand-alone article on a subtopic. WP:SPINOUT is pretty vague about the appropriate circumstances, and makes no reference to a difference in support requirements. Nothing there is germane to allowing material in a longer article to be sourced to a primary source, but refusing that same information to appear in a stand-alone article simply and solely because its support is a primary source. I understand that arrangement is present practice, but it makes no sense.

Yes, the stand-alone may be inadvisable for many reasons, but if the article falls into the classification of a "familiarization article", primarily relying upon primary sources is not one of them. You also raise the prospect that the stand-alone is not sufficiently notable to become a separate article. Here also, if the stand-alone is of "familiarization article" type, the notability issue should be decided based upon various criteria we have touched upon, and not entirely upon whether or not it relies upon a secondary source. Brews ohare (talk) 21:55, 10 February 2012 (UTC)

I imagine "Chad (Governmental structure)" can be referred to a secondary source. Assuming that were possible, the secondary source would say something like "Chad has a constitution. (See Constitution of Chad)", which could be equally authoritatively directly referenced as a primary source. (Actually, in this example, I don't think Constitution of Chad is a primary source, as that is probably in French.) What is aimed at is removal of the silliness of finding source S that simply quotes primary source P, when source P can be cited directly with equal authority. Brews ohare (talk) 21:55, 10 February 2012 (UTC)

First of all, we're still not talking about practical "familiarization" articles - these are not navigational in any shape or form. These are articles in mainspace, and need to meet content and inclusion guidelines. So be aware, you've still not identified what exceptions would be needed here.--MASEM (t) 22:57, 10 February 2012 (UTC)

Yes. Brews ohare (talk) 16:17, 11 February 2012 (UTC)

Now, let's compare the two cases "the organization of CZ" and "the government of Chad". Barring anything else, as long as we can source these via primary sources that meet WP:V, such could easily be covered in the respective parent articles (CZ and Chad - both are clearly notable topics). The question you're asking is why is there is a issue if these were separate articles.

If we talk about the "government of Chad", that is a clearly notable topic of its own. It is assuredly covered in textbooks within the country and by political science/global government textbooks published elsewhere, at a minimum (I don't have time to do a sources run, but I hope this is an accepted point). Within the article, however, the "best" resource is the primary source - the country's own constitution and laws that describe the organization of the government. So we have a topic that is notable where there do exist secondary sources, but readily sourced to the primary. In such a case, the lead for the article simply needs to establish the notability (and arguable, the notability is so obvious that most people won't flinch at the lack of secondary sources since they are assured it would exist), and the rest of the article could be on primary sources. In the context of this discussion, the resulting article may only include primary sources, but we know that secondary sources exist that talk about the topic "The government of Chad". Note - this is not the issue with the proposal, because this is not the case where only primary sources exist.--MASEM (t) 22:57, 10 February 2012 (UTC)

So, is it correct to summarize as follows? In a longer article, a primary source may be preferable to a secondary source for some matters (just which could be made very clear if desired) and a spin-off of this material to a stand-alone article is OK using the same primary sources provided (i) notability can be established somehow, or failing that, (ii) secondary sources are known to exist, even if they actually are not used. Brews ohare (talk) 16:17, 11 February 2012 (UTC)

Now, take the case of "the organization of CZ". We know CZ is a notable website, but is it's organizational structure notable? Certainly there's no way I'd take the word of anyone on that, like I would with the government structure of a major country. Let's presume that there are no secondary sources that talk about "the organization of CZ". (about "CZ" yes, but not the organization). This article would then have to be sourced to the primary work, the pages of CZ itself. This is the article that fails the policy as only primary sources exist for the topic; this article thus should not exist as a stand-alone article, but still can be discussed within the body of the "CZ" article.

If it were the case that there was significant coverage by secondary sources about CZ's organization, that makes it a different story: now we can include that secondary sourcing as the lead for the article (and here, unlike the Chad case, it would be expected to see that), and then have the rest of the article using primary sourcing. Mind you, the appropriateness of splitting this article off may be questioned: if both the CZ and the "organization of CZ" articles are short, then it is better to keep them as a single article; splitting off should only be done when the topic is notable and size issues start coming into place.--MASEM (t) 22:57, 10 February 2012 (UTC)

Here we discuss spin-off of material where notability cannot be established, or is debatable. For some reason, the criterion for notability here is taken to be the existence of secondary sources. The type of case considered is where only primary sources are available, and it is pointed out that such a spin-out is not presently admissible. There is no doubt that this is present policy.

What is not discussed at all is the subject of circumstances under which this policy should be relaxed. In particular, there are articles that can be sourced only through primary sources, where these primary sources are the best support imaginable for the topic at hand, and where a secondary source, even if it existed, would simply parrot the primary source. The only valid issue here is whether notability for the article can be established without citing a secondary source to establish the topic is worth talking about.

That is possible in some instances (you have provided a couple of them already). So I'd guess the subject of this conversation should include a discussion of how to establish notability in the absence of secondary sources, and possibly even the topic of just how a secondary source establishes notability when it is available. I'd guess the latter is a grossly misused mechanism in practice. Brews ohare (talk) 16:17, 11 February 2012 (UTC)

So to stress this point: the issue is between articles where the only sources in the world on the subject are primary, and articles where at the present time only primary sources are included but secondary sources are known to exist. The latter case does not fail OR policy/N guideline (though really those secondary sources should be added to prevent the issue), but the former one does and should be merged/redirected/deleted to a larger topic. Remember: once within the article of a larger topic, a smaller topic can be discussed to appropriate due depth from primary sources without question. --MASEM (t) 22:57, 10 February 2012 (UTC)

My view of this matter is that the imbedded section achieves notability in the larger article because it is part of a larger discussion or presentation and serves to illustrate some aspect of the main topic. As such, the notability of the section does not apply to the section itself when it is divorced from the larger topic. I believe this is a paraphrase of what you think yourself.

However, I'm not interested here in any kind of subject matter. The material I'm interested in is simple factual description, for instance, the structure of an organization, and not material involving evaluation, interpretation, or synthesis. For such material, primary sources are the most reliable support. The notability of such an article is not established by the significance of its views (it doesn't express any) and so does not require a secondary source to support its views. Its notability could be established by other means, for example, one might claim the structure of an organization was notable because the organization itself is notable. Perhaps, what is needed is a secondary source saying, for example, "IBM is a billion dollar company [Source: Barrons], with the following structure", or "Citizendium is an on-line encyclopedia designed differently than Wikipedia[Source: USA Today]. What follows is an outline of its structure", or "Rabies in dogs occurs [Veternary Science] and municipalities have regulation regarding this matter, as outlined below." These topic sentences establish context, perhaps, but not a thesis that is the article content. How would that fly? How would you adjust this approach? Brews ohare (talk) 16:17, 11 February 2012 (UTC)

You are in the right direction that you should establish the notability of the topic at the start of the article with secondary sources (or make sure that you have identified them on the talk page) to meet both WP:OR (that the topic has been covered and thus the rest is not undue) and WP:N (as general inclusion guidance); once established the rest of the article could be all primary sources though I think there may be some caution about going too far overboard on details from primary sources only, but that's not the issue at this point. But we you have to recognize is what topic that article now is about. If I broke out "Government structure of Chad" from "Chad", the topic I have to show notability is "Government structure of Chad". Similarly, breaking out "Organization of Citizendium" from "Citizendium", I have to show that "Organization of Citizendium" is a notable topic, even knowing that "Citizendium" itself is notable already. If you can't show that, you shouldn't break out the article , but discussion from primary sources is still fine in the original parent article. --MASEM (t) 16:36, 11 February 2012 (UTC)

Masem: I've raised the point as yet not addressed in detail of just how to establish notability. Of course, if you are trying to establish a thesis (an evaluation or interpretation or whatnot) a secondary source will provide support that the thesis is of sufficient note to have received some attention. But what are the other ways to establish notability? You suggested that something like "District attorney" was a notable topic because it was part of the court system or system of justice, and that is in itself notable. Would it suffice for an article like "State regulations regarding district attorneys" to simply have a secondary reference noting what the term "district attorney" means, and then proceed with description based upon primary sources? How could one formulate some suitable approach to deal with such things in general, apart from this particular example? Brews ohare (talk) 02:31, 12 February 2012 (UTC)

No, that would not be good enough. You're basically saying that any sub-topic of a larger topic can be given an article and talked about in depth even if the sub-topic is non-notable because the larger topic is notable. That's basically the idea of inherited notability, and that is not appropriate on WP. --MASEM (t) 12:53, 12 February 2012 (UTC)

No, the question I am raising is that of how to establish notability without invoking a secondary source. You noted earlier that you thought articles that could be classified a "familiarization articles" do exist, and are acceptable. Would you say they are exempt from notability requirements, or is their notability determined without the use of secondary sources? Brews ohare (talk) 13:59, 12 February 2012 (UTC)

There is no way to establish notability without significant coverage by secondary sources, otherwise we are putting undue weight and bias on primary sources. Now, yes, navigational articles like disambiguation pages or outline do not need to show notability. What that problem is is that you're using this term "familiarization article" and applying it to things that are definitely not navigational pages (eg the suggestion of a page about "State requirements for DAs"). As I've asked you before, if you provide examples (existing ones at best) that you think are "familiarization articles", we call tell you if they would have to qualify for notability or not. --MASEM (t) 14:08, 12 February 2012 (UTC)

Masem: So to state your position in my own words: There is no way to establish notability other than using a secondary source. However, some acceptable articles do not need to be notable, and the guidelines to identify these exceptions are sufficiently obscure or subtle that no formal policy can be written to identify them. Rather, in case an article is challenged as not exempt, the matter should be referred for adjudication to Masem, who will make the determination as to whether an exception is warranted. Brews ohare (talk) 15:00, 12 February 2012 (UTC)

This is well-established practice, and takes no guesswork as to what is exempted. Navigational articles like disamb pages or outlines are exempt from generally all other content and inclusion guidelines for articles, because they are strictly that, navigation aids. Any articles outside that are assumed to be normal encyclopedic articles and thus must mean all content and inclusion guidelines need to be met. The problem is that you see to want to push normal content articles as a "familiarization" article to make them exempt from content and inclusion guidelines, which is against current policy. --MASEM (t) 15:06, 12 February 2012 (UTC)

No, that is not my intention. I guess the closest that present guidelines come to this matter is Notability: Lists. That guideline section already admits to confusion on this matter. It makes the recommendation: "Editors are still urged to demonstrate list notability via the grouping itself before creating stand-alone lists"; which is completely unintelligible to me.

As you may recollect, I am not focused here on "normal content" but upon content that is purely descriptive, contains no evaluation, interpretation, or opinion, and is based primarily upon primary sources. Of course, a navigation page like Outline of chemistry is in this category, and kind of resembles a knol collection. That is, a guidance page governing material internal to WP, which is useful to a reader searching for what there is on WP about the topic. To a degree it is a very restricted pre-programmed Google search like this. So I raise this question: why wouldn't one assume that such pages are useful also for outside information sources? Brews ohare (talk) 15:31, 12 February 2012 (UTC)

Because basically, even if the information is fully sourceable to primary and thus verified, if no one else has taken note of it, it is undue weight for us to cover it in depth on Wikipedia. We cannot create an importance for a topic if there is no importance given to it in sources. --MASEM (t) 16:51, 12 February 2012 (UTC)

Another possible candidate for a "familiarization" article is United States Attorney. It does have a few secondary sources, but it appears to be debatable whether any of these establish notability, and it raises the issue of whether it is primarily based upon secondary sources: it doesn't look that way. Yet, I'd say the article has a purpose and should be on WP. Brews ohare (talk) 15:59, 10 February 2012 (UTC)

That's not what most people would consider a navigational article. "United States Attorney" is a notable topic since they're a critical part of the national court system, and clearly backed by secondary sources that establish its notability. --MASEM (t) 16:05, 10 February 2012 (UTC)

The criterion "they're a critical part of the national court system" as a basis for notability is an example of a basis different than requiring a secondary source. It is the major reason for considering this article notable, and would suffice by itself. Most of the secondary sources in this article refer to some particular minor point. So this article seems to fit the bill as an article denied existence by WP:Notability WP:V and WP:Primary that should be on WP nonetheless. It is not primarily based upon secondary sources.

Maybe you don't like the name "familiarization article" (I don't either) but "a rose by any other name..." Brews ohare (talk) 16:15, 10 February 2012 (UTC)

It's not a navigation, familiarization or whatever other name you want to give to call these articles "special". Unites States Attorney is a normal, everyday mainspace article that shows its notability through secondary sources. Some sources are primary, which are fine, but the topic is notable to secondary ones. So this is still a bad example to start from. --MASEM (t) 16:30, 10 February 2012 (UTC)

Masem: It is suitable in the sense of being an article whose "notability" is independent of the presence of secondary sources. That shows that notability can be established in some cases without secondary sources, perhaps on a case-by-case basis?

Are you suggesting that this "bad example" indicates that no category like "familiarization article" is needed? Brews ohare (talk) 16:34, 10 February 2012 (UTC)

No, I'm not saying there's no such thing as a "familiarization article". Just that United States Attorney is not the type of article that would be considered as such. It is a notable topic (you have to consider what full range of sources exist, not just those that are presently listed in the article), and not used for navigation. Just so you know what I consider as such, Outline of chemistry is one example of a navigational, or familiarizational article. --MASEM (t) 16:48, 10 February 2012 (UTC)

There appears to be many articles in the area of officialdom that need some attention: District attorney, Prosecutor, Crown attorney... and may end up being familiarization pages. Brews ohare (talk) 16:15, 10 February 2012 (UTC)

I suppose they could end up the way you envision... the question is: should they end up that way? I would say No, they should not. And if they did end up as you envision, I strongly suspect people would object and re-write them back into normal articles (hopefully citing reliable secondary sources).

I think you need to answer a further question here... WHY would we want to have "familiarization articles"... what would be their purpose? Blueboar (talk) 17:53, 10 February 2012 (UTC)

Of course, an article evolves over time, and one might hope that an article based upon primary sources might become a regular article, for example, if a book happens to be written one day that covers the topic. In the meantime, the article can sit in a different category. At the moment, if any challenge is made to an article based primarily (whatever that is taken to mean) upon primary sources, it must be removed. That means topics like the structure of organizations that have no secondary coverage cannot be written. Of course, there are such articles on WP that escaped challenge, but if there is any opposition (such as arose in trying to include WP:Formal organization in the article Wikipedia) the article is denied to Main space. Brews ohare (talk) 18:07, 10 February 2012 (UTC)

It is clear that good reasons for rejecting an article should not be ignored, but rejection of a "familiarization article", defined as defined above, would not happen simply and solely because its sources are primary. Brews ohare (talk) 18:10, 10 February 2012 (UTC)

Brews, the place to discuss the "structure of an organization" is within the article on the organization itself. The structure of the organization is very unlikely to be considered a notable topic on its own (and in the rare cases where it might be, due to some unique aspect of that structure, we would determine its notabiliy by having secondary sources discuss that structure).

This actually gets us back on track as far as the OR policy goes... I would absolutely agree that if we were writing a section on the structure of an organization (in larger article on the organization itself), it would be fine to base that section primarily on an a Primary source... after all, the most reliable (ie best) source possible for the information is highly likely to be a primary internal document from that organization. We would, of course, have to be very careful not to slip into analysis or draw any conclusions... but as long as what we are writing is purely descriptive, a primary source is fine. But... we are talking about a section. The article on the organization, as a whole, would still need to be mostly based on secondary sources. Blueboar (talk) 23:36, 10 February 2012 (UTC)

Blueboar: Do you think the present policy as written is satisfactory regarding OKing primary sources for sections of articles? I am doubtful. Brews ohare (talk) 02:18, 12 February 2012 (UTC)

I think the present policy is OK as far as permitting the use of primary sources (it explicitly says you are allowed to use primary sources, but cautions that there are limitations. It does a good job of explaining when it is not appropriate to cite primary sources (statements of analysis, interpretation, conclusions, etc.), but I have long thought it could do a better job of explaining: 1) when it is appropriate to cite primary sources, and 2) how to do so in those situations. Blueboar (talk) 02:33, 12 February 2012 (UTC)

Blueboar: What is your take on the list of articles on WP provided here? How is notability established for these separate articles? It seems probable that the way to achieve notability is largely established by forming a "project" like WikiProject Alaska that includes the document. Unfortunately U.S. Attorney for the Eastern District of Kentucky hasn't evolved such a group as yet. Brews ohare (talk) 15:51, 12 February 2012 (UTC)

I opened a thread on Talk:Notability where you may prefer to take up this matter. Brews ohare (talk) 16:57, 12 February 2012 (UTC)

Can someone break and possibly summarize this list and what currently supported options are? I'm having trouble drawing a conclusion from this long debate.Gsonnenf (talk) 23:08, 19 February 2012 (UTC)

Based on what ground the statement "XXX is a term" can be made in WP articles?

From time to time, I encounter the WP articles that start with the statements of that type:

"XXX YYY is a term used for XXX by YYY." (The most recent example can be found there: Soviet occupations is a term used for military occupations by the Soviet Union...).

As a rule, a source used for such a statement contains just a combination of XXX (as a subject) and YYY (as a predicate), and contains no definition of such new term. For instance, the sources 1-3 from the above example contained no definition of "Soviet occupation" as some separate term used for for description of some separate phenomenon. They just tell about occupation by the Soviet Union. In my opinion, taking into account that I have been a witness of numerous attempts to create new "terms" using the same approach, it would be useful to describe some of them as an example of WP:OR.--Paul Siebert (talk) 02:54, 9 February 2012 (UTC)

Hmmm... this could be OR. However, I think it is an OR that could probably be fixed by adding attribution. Change the sentence to: Soviet occupations is a term used by historians to describe various military occupations by the Soviet Union..." (followed by citations showing historians using the term to describe those occupations) and it would not be OR. Blueboar (talk) 03:32, 9 February 2012 (UTC)

In the case at hand, and noting ArbCom decisions and dicta thereon, the lede is proper as is. A plenitude of sources are given in the articles, and thus nitpicking over grammar is non-utile. I suggest you read the ArbCom discussions on all of this "stuff." Collect (talk) 14:07, 9 February 2012 (UTC)

I am not sure I understand what concrete ArbCom decision do you mean (maybe you provide us with the link?), however, none of the sources given in the article define the term "Soviet occupation", which is supposed to be a specific type of occupation to deserve a separate term.--Paul Siebert (talk) 15:35, 9 February 2012 (UTC)

I have come across people making up terms and they are obviously OR but others are not. In the case of soviet occupation it just sounds like clumsy language and unnecessary. We are allowed to use descriptive names for article titles if there is a clear topic and no agreed name or term. Dmcq (talk) 14:19, 9 February 2012 (UTC)

We are definitely allowed to use descriptive names, however, we are not allowed to pretend they represent some separate terms. In connection to that, I propose to describe the difference between descriptive names and the terms. Let me explain that using the following (totally imaginary) example. Consider two statements:

1. "Medieval negativism is a term used by historian to describe the mainstream philosophical discourse in Middle Age Europe.¹"
2. "During Middle Ages, the European philosophical thought was dominated by negativism.¹"

(Again, the example is totally artificial, so, please, ignore the factual side of the sentences #1-2). Depending on what the source 1 says, the statement 1 may be allowed or not. If the source contains an explicit definition of "Medieval negativism" as a term, the statement 1 is acceptable. However, if it just contains "medieval" (a subject) and "negativism" (a predicate), and does not define the term, the statement 1 fails WP:NOR. Since I have been a witness of several attempts to introduce such new "terms", the above situation is not unique, so we probably need to explain the difference between the terms and the descriptive names either on the policy or guidelines page.--Paul Siebert (talk) 15:35, 9 February 2012 (UTC)

Paul, your basic interpretation of how this policy works is accurate... the determination as to whether a specific statement is OR or not can indeed depend on the exact language used in the cited sources. If historians actually use a specific term to describe something, then it is not OR to say so. If they don't actually use the term, then it would be OR to say that they do.

However, I don't think we need to change the policy to resolve your concern. The policy already requires that sources "directly support" what is said in an article. There is no need to explain this provision in terms of every instance where it might apply. Doing so is what we call "instruction creep". Blueboar (talk) 17:00, 9 February 2012 (UTC)

That said... I have looked through both the edit history of the Soviet occupation article and your personal edit history... I suspect that your real concern here is WP:NPOV rather than WP:OR If so, I suggest you address that issue directly. As it is, there is an appearance (which may or may not be accurate) that you are trying to wikilawyer about OR in order to "win" a NPOV dispute. If I am misreading things, I apologize... but that is how it appears to an outsider. Blueboar (talk) 18:12, 9 February 2012 (UTC)

By writing that my primary concern is NPOV you were right. By writing that I am trying to wikilawyering to win some dispute you were wrong. I simply found some typical example of creation of new "terms", and I think we need to clarify the policy to make such violations of NOR impossible.--Paul Siebert (talk) 20:18, 9 February 2012 (UTC)

With few exceptions, "XXX is a term" is usually just an example of weak writing style, not a policy violation.

Importantly, the lead is supposed to summarize the article, not provide dictionary definitions. So the best approach in such situations (using Blueboar's example) is usually something like

Medieval negativism was the mainstream philosophical discourse in Middle Age Europe, which was characterized by... WhatamIdoing (talk) 18:53, 9 February 2012 (UTC)

Correct. That is definitely better. However, my argument is that someone may argue that the difference between ##1 and 2 (or your test) is just stylistic one, whereas the difference is more profound: the # 1 is a violation of NOR, and that should be explained clearly.--Paul Siebert (talk) 20:18, 9 February 2012 (UTC)

Comment: My two cents!- OR is a rule generally employed very skillfully. The object here seems simple: if a citation can be found that says 'X is known for' or 'X was defined by' (which gives us a 'term' applying to 'X'), I don't see that as OR. The Medieval example is simply too abstract and niggling for the general purpose of the query. I.e., my understanding is that Wikipedia is not a dictionary. When you speak of 'terms', aren't we trespassing on the dictionary's field? So, what I say is if a citation is found that is clear enough and not a matter of definitions or semantics that lay outside Wikipedia's scope, it is acceptable; otherwise it is OR.--Djathinkimacowboy 07:50, 10 February 2012 (UTC)

I do not propose to consider a situation when such a citation is found. I focus on the opposite case: when editors, based on handful books that use some subject and predicate together, create a new "term". In my opinion, that should be explicitly prohibited: no sourced definition - no term.--Paul Siebert (talk) 15:08, 10 February 2012 (UTC)

What this seems to come down to is: While we are allowed to invent our own phrasings and terminology when creating a descriptive title (with the caveat that if there are existing, commonly used phrasings and terminology for the topic, we should use those instead)... it is not always appropriate to describe our invented phrasings and terminology as being "a term" in the text of the article. ie - We should not use the word "term" unless a source uses the exact phrasing/terminology as a term.

I would say this is a very narrow and nit-picking interpretation of WP:OR (bordering on being Wikilawyerish)... but I have to agree that it is OR. We require that sources directly support the text. In the situations Paul is describing they do not (they indirectly support the text). The solution is not to prohibit the phrasing/terminology... the solution is to re-write the opening sentence to avoid using the word "term". Blueboar (talk) 15:40, 10 February 2012 (UTC)

I agree it seems to be more of a case of clumsy language when some editors attempt to apply MOS:BOLDTITLE, and the solution is as you suggest and easily applied[1]. --Nug (talk) 15:49, 10 February 2012 (UTC)

In re-reading the proposition as originally stated by Paul Siebert, I see the problem as a simple one: no editor can or should infer/invent a new term solely based on two terms seen used together in a citation. In other words, as another wise commentator said, if it isn't verifiable in a cited source, it is OR. I recall being whipped nearly to death once when I tried to describe something. I used the term 'ironically' and got the lash for it. I hope that comparison example is clear. Let the reader infer whatever the cited material tells, and no more.--Djathinkimacowboy 01:40, 16 February 2012 (UTC)

• My handy rule for history articles:
  • Where the scholarly discourse is dominated such that only a scholarly FRINGE disagree, then the term is the thing. "World war II was a conflict between the Allies and the Axis."
  • Where the scholarly discourse is internally divided, it is a term, "The Great Patriotic War is the term used in Soviet historiography to describe the front in World war II between Germany and its European allies and the Soviet Union."
  • Where a scholarly discourse is a minority position, it is a term, "Post-fordism is one of a number of terms to describe the structure of employment and the economy in the advanced West after 1970."
  • Where a scholarly discourse is FRINGEy, it is a term, "Within Novocastrian provincialist gutter murmurings, "Newcastle Hardcore" is that particular style of Hardcore techno developed in Newcastle, NSW between 1990 and the mid 2000s."
  • Where it ain't scholarly and it is FRINGE, it shouldn't be used.
• Of course, I strongly suspect (but feel insufficiently read to claim) that "Holodomor" ought to discuss Ukrainian remembrance, whereas the Ukrainian Famine article should be about the object. Consider regarding your own views, and with salt. I think my schema is internally consistent, and reflects the WEIGHT policy. Fifelfoo (talk) 03:09, 16 February 2012 (UTC)

Clarification of what constitutes a primary or secondary source

The description of primary and secondary sources is not as clear as it might be on this policy page. At the moment, the policy says under Secondary sources:

"For example, a review article that analyzes research papers in a field is a secondary source for the research."

That is very clear. What is less clear is whether this statement implies that the research papers themselves are primary sources. A footnote is intended to clarify this matter:

"The Ithaca College Library compares research articles to review articles. Be aware that either type of article can be both a primary and secondary source, although research articles tend to be more useful as primary sources and review articles as secondary sources."

I find that this footnote serves only to muddy the waters. It can only lead to debate over whether a particular journal article falls under multiple categories and which category applies in what instance. I'd suggest it be revised to say the following:

"Research articles are considered primary sources and review articles as secondary sources."

That is crystal clear, and does not leave the door ajar.

The reason journal articles presently are allowed as secondary sources is most probably related to journals being peer reviewed. That review process may be adequate for some technical journals like Physical Review, but even there articles may not reflect a broad consensus of the technical community. In fact, they report original research that by definition may not be conventional in its approach. Unless WP wishes to go into the unlikely business of ranking journals, the simplest policy is to regard them all as primary sources.

Another footnote in WP:NOR is:

"Duke University Libraries offers this definition: "A primary source is a first-hand account of an event. Primary sources may include newspaper articles, letters, diaries, interviews, laws, reports of government commissions, and many other types of documents."

Placing this information in a footnote in this manner suggests that this statement is to inform judgment, and is not an integral part of the WP policy on this matter. In the interest of clarity and to ensure objectivity of article content, I'd suggest that wording of this sort be included explicitly in WP:NOR, for instance, with the statement:

"Newspaper articles, letters, diaries, interviews, laws, reports of government commissions, journal articles, and many other types of documents are considered by WP to be primary sources."

Brews ohare (talk) 18:49, 15 February 2012 (UTC)

Comments on what constitutes a primary or secondary source

• Comment: The situation really is muddy. If a policy or guideline requires a precise definition of primary and secondary sources in order to be interpreted, the policy or guideline does not reflect the real world and needs to be changed. Jc3s5h (talk) 18:52, 15 February 2012 (UTC)

Jc3s5h: If I understand your comment, if the real world is confusing, a guideline must be confusing too. If it isn't as muddy as the real world, the guideline must be made murkier. Have I got your meaning? Brews ohare (talk) 18:57, 15 February 2012 (UTC)

No. Since the distinction between primary source and secondary source is murky, any policy based on this distinction will be murky too. The alternative is to base the policy on something else; something not so murky. This may not be possible, but then again a way might be found. Jc3s5h (talk) 19:03, 15 February 2012 (UTC)

• Comment: It seems to me that this probably cannot be resolved in a really satisfying manner, as use of the terms primary and secondary sources depends on the concrete context and field. The keyword in list of publication forms in the Duke definitions is "may". I see no way to classify news paper articles, journal articles or commission reports principally as primary sources, that would be nonsensical. For example a commission report that simply collects facts/information is certainly a primary source, however a commission report by scholars that assesses and reviews facts,information or primary sources in general is no different from a a book or journal article reviewing primary sources. In the context of history subjects you may consider all those listed as principally (historic) primary sources, but that works only in this context (due to the time difference to current publications on the subject). If you move to other fields that approach fails. Also I'd like to note that we should avoid getting hung up on (possibly unachievable) formal distinction between primary, secondary and tertiary. It is much more important that editors focus on the quality of the content of a source and its reliability and reputation, that is what matters primarily and not the formal distinction between primary, secondary and tertiary.--Kmhkmh (talk) 19:16, 15 February 2012 (UTC)

It is certainly true that there are gray areas. You suggest that some government reports are reviews, and so are some journal articles. I don't see much problem in identifying these as secondary sources. On the other hand, many journal papers begin with a summary of prior work, that is in a sense a review. I believe excluding such introductory sections from consideration as a secondary source is warranted, especially as such introductions are generally intended to cast the subject in a light to reflect favorably upon the original research that is the purpose of the journal article.

Can you elaborate upon your reservations? Brews ohare (talk) 19:26, 15 February 2012 (UTC)

As Kmhkmh notes, when they're older than about 10–50 years, newspaper articles become primary for history. In Medical Science, research articles are primary, whereas research reviews are secondary. A number of commission reports are primary, where the commission is an involved party, consider the report of the Winter Soldier Investigation. The provision of a brief literature review in historical journal articles is considered to be a secondary source and is privileged especially if the whole work is a literature review for weighting and structure in history articles. Context is all here. Fifelfoo (talk) 22:47, 15 February 2012 (UTC)

• Comment: Oppose. The proposed edit " Newspaper articles, letters, diaries, interviews, laws, reports of government commissions, journal articles, and many other types of documents are considered by WP to be primary sources." is a blatant oversimplification. Below is my analysis:

1. "Newspaper articles." All newspaper articles? Whereas some news reports, where the facts have been presented without any interpretation are, probably, the primary sources, the analytic articles are definitely not primary.
2. "Letters." Whose letters to whom and about what?
3. "Interviews" What about an interview of some WWII historian about his work on the history of the attack of Pearl Harbor? Is it a primary source?
4. "Laws" Yes. It is a primary source.
5. "reports of government commissions" Sometimes, such reports contain a comprehensive analysis of some event (e.g., of some catastrophe), its causes, consequences, etc. That fully meets a definition of a secondary source ("are second-hand accounts, generally at least one step removed from an event").
6. "journal articles" In exact sciences, journal articles are divided onto research articles and reviews. The latter are definitely secondary sources. The former contain the author's own data, their discussion, and more or less comprehensive overview of what have been done by others in this field. Therefore, depending on the context, research articles are either primary or secondary sources. With regard to history articles, each of them contain a huge amount of references to numerous archival documents, letters, research articles, books and monographs, which definitely makes it a secondary source.
7. "and many other types of documents" That addition makes the whole sentence useless. Which concrete other documents?

And finally, let me remind you that the PSTS division exists not for the sake of itself, but as a part of our NOR policy. We cannot interpret raw data, historical documents, laws, etc by ourselves, therefore, the sources that contain no such interpretation should be used with great cautions in Wikipedia. However, the sources where such information have been interpreted by professionals (research/hisotry articles, reviews, letters of one scientist to another, philosopher's diary, etc) is safe to use.--Paul Siebert (talk) 19:28, 15 February 2012 (UTC)

Paul Siebert: It appears that you have your particular definition of primary sources, namely "raw data, historical documents, laws, etc." Perhaps you could express these ideas in a form that would clarify WP:NOR? Brews ohare (talk) 19:35, 15 February 2012 (UTC)

No. I follow the definition of a secondary source as "second-hand accounts, generally at least one step removed from an event". The history article dealing with historical documents (primary sources) is definitely at least one step removed from the event it describes. The interview with a present-days physicist about the Manhattan project is also at least one step removed from the event, and so on. I do not propose my own comprehensive definition of primary sources, I just outline those sources that are always primary. Other sources may be primary or secondary, depending on a context.

And, again, my major point is that we always have to keep in mind that the real reason for the PSTS division is to avoid original research, which mean to avoid making amateurish interpretations of raw data. If some source contains an opinion of some professional, it is safe to use it in Wikipedia, and that is a major point of the NOR policy, in my opinion. Everything else should be built on based that.--Paul Siebert (talk) 19:50, 15 February 2012 (UTC)

Paul, I'll take the opportunity to look at your specific points above:

1 "Newspaper articles." All newspaper articles? Whereas some news reports, where the facts have been presented without any interpretation are, probably, the primary sources, the analytic articles are definitely not primary.

I personally would not accept an interpretive newspaper article as a secondary source. These articles are labeled as opinion pieces and that is what they are. Their only value as a source is as a primary source regarding what the analyst thinks.

Actually even that is imho somewhat fuzzy since, we need to distinguish between "interpretation" as a scholarly (well researched) assessment and interpretation as a mere (political) journalistic opinion piece.--Kmhkmh (talk) 22:35, 15 February 2012 (UTC)

Additionally, the assertion that all interpretive articles are "labeled as opinion pieces" is simply false. If a periodical runs a long series analyzing an issue, e.g., comparing current proposals for tax reform, that's not an opinion piece. WhatamIdoing (talk) 05:06, 23 February 2012 (UTC)

2 "Letters." Whose letters to whom and about what?

Any letter is a personal observation, or a hearsay version of other peoples', and is not a secondary source.

3 "Interviews" What about an interview of some WWII historian about his work on the history of the attack of Pearl Harbor? Is it a primary source?

Yes, it is a primary source regarding his views of his work. On the other hand, the historian's work itself is treated on its own merits as to whether it is primary or secondary in nature.

4 "Laws" Yes. It is a primary source.

5 "reports of government commissions" Sometimes, such reports contain a comprehensive analysis of some event (e.g., of some catastrophe), its causes, consequences, etc. That fully meets a definition of a secondary source ("are second-hand accounts, generally at least one step removed from an event").

Yes, and where these reports are secondary sources it is quite easy to establish. For instance, if they assess the efficacy of government agencies, they are secondary sources, although possibly not reliable.

6 "journal articles" In exact sciences, journal articles are divided onto research articles and reviews. The latter are definitely secondary sources. The former contain the author's own data, their discussion, and more or less comprehensive overview of what have been done by others in this field. Therefore, depending on the context, research articles are either primary or secondary sources. With regard to history articles, each of them contain a huge amount of references to numerous archival documents, letters, research articles, books and monographs, which definitely makes it a secondary source.

Overviews in reports of original research may be accurate, or they may cast prior work to showcase the following original work. As such, these overviews should be viewed as primary sources about what the authors think. In eliminating all but review articles from consideration as a secondary source, not much is lost.

7 "and many other types of documents" That addition makes the whole sentence useless. Which concrete other documents?

Good point; this language was inadvisably carried over from the original source.

Brews ohare (talk) 20:02, 15 February 2012 (UTC)

1. May I ask you, what is the reason for division of the sources onto primary and secondary?

2. As a rule, the reviews and research articles are being written by the same authors (I myself have authored both research articles and reviews in my real life). In connection to that, consider the following situation (i): an author A reviewed the works authored by X, Y, and Z, and published that in the review article; (ii) the same author reviewed the same works and published the same material in the introduction to his own research article. What makes the former source secondary for X, Y, and Z, and why the same text in the latter source is primary?

3. Sometimes, the monographs are the collections (with small additions) of the research articles written previously by the same author. What makes the book (monograph) containing several articles (e.g. as chapters) a secondary source, and why the same research articles published in some scholarly journal separately are primary sources?--Paul Siebert (talk) 20:30, 15 February 2012 (UTC)

Hi Paul: I confess that this topic is more subtle than I had first thought. From a WP stance, the role of a primary source is to support some statement of fact, while a secondary source has a more involved role. The secondary source can be used to establish WP:Notability and a primary source cannot. Personally, I find that nuts.

It may be that on WP there is no other point in separating the two types of source, and it is just nonsense. The only point of importance is that contributors to WP should not intrude their own opinions, and that can be established with any type of source.

So our conversation here is leading me to the view that all discussion of primary and secondary types of sources should be eliminated from WP and WP:Notability should be rewritten entirely to establish different means to establish whether an article is worth appearing on WP. Maybe any topic for an article should be allowed, subject to deletion if they have too low a traffic count? Brews ohare (talk) 20:54, 15 February 2012 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ No. This is silly and needs to stop. Its tendency is towards disruption. Fifelfoo (talk) 21:23, 15 February 2012 (UTC)

• Comment:

I note that this discussion in this section is a few steps removed from the previous section, after Brews ohare has tried to broach similar issues at other boards. --MASEM (t) 21:03, 15 February 2012 (UTC)

• Comment: Agree with the above posts by both Masem and Fifelfoo.--Djathinkimacowboy 01:31, 16 February 2012 (UTC)

"Maybe any topic for an article should be allowed, subject to deletion if they have too low a traffic count?". No, maybe not. If your real aim here is to eliminate WP:N and replace it with WP:INTERESTING then please open an RfC to this effect and see it snow-closed as Oppose. Fences&Windows 21:33, 22 February 2012 (UTC)