Wikipedia talk:No original research/Archive 42

This is an archive of past discussions about Wikipedia:No original research. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 35

←

Archive 40

Introductory sections of papers

This text was added and removed from the policy:

"A scientific paper that presents original results often includes a section, typically the introduction, where related work is reviewed. This may constitute a suitable secondary source in the absence of dedicated review papers in that area."

The edit summary for removal was "You misunderstand the concept of Original Research". Now, I am very comfortable with the WP concept of original research (I've been participating on this page for quite a long time). I don't see how the concept of avoiding original research affects our use of an introductory section of a paper as a source. Moreover, the language quoted above seems like a perfectly good description of the actual practice I have seen employed in various articles. — Carl (CBM · talk) 20:08, 13 January 2009 (UTC)

Perhaps my edit summary is not worded correctly... I have explained my revert more fully below (we cross posted). Blueboar (talk) 20:17, 13 January 2009 (UTC)

PSTS revised

The PSTS section offers advice on the use of types of sources prior to defining those types. It defines the three source types backwards (a failed experiment IMO). It contains repetitive text and flabby writing. It contains material that does not belong in this policy page (such as comments on reliability). Finally, the primary sources section contains some examples that are typically unpublished and so irrelevant for our purposes.

I present a revised version that begins with a general rule, three definitions with examples, and then discussion on the use of these types as sources on WP.

The grouping of sources into primary, secondary or tertiary is governed by their originality and their proximity to the origin.[1][2][3]

Primary sources are original material, original thinking that has not been previously published, or first-hand descriptions of contemporary events. Examples include artefacts; recordings; diaries; interviews; scientific journal articles presenting original research; letters; first-hand newspaper accounts; patents; government records; photographs; religious texts; and creative works such as art, architecture, literature and music.

Secondary sources describe, comment on, discuss, interpret, analyse and evaluate primary sources; they are one step removed from the origin; they are typically written later, by a third party. Examples include biographies; commentaries; scientific literature reviews; humanities journal articles; second-hand newspaper discussion; specialist textbooks and monographs.

Tertiary sources collect and distil the information contained in primary and secondary sources. Examples include encyclopaedias, almanacs, guidebooks, university textbooks, bibliographies and catalogues.

Within one source document may be both primary and secondary material. Categorisation also depends on what facts are drawn from a source. For example, a magazine article on a scientific discovery (recently published in a scientific journal) is a secondary source on that discovery, but a primary source on what the media are saying about the discovery.[4]

Wikipedia articles should rely mainly on secondary sources and, to a lesser extent, on tertiary sources. All interpretive claims, analyses, or synthetic claims about primary sources must be referenced to a secondary source. Primary sources may be used in Wikipedia, but only with care, because it is easy to misuse them. A primary source may be used only to make descriptive claims, the accuracy of which is verifiable by a reasonable, educated person without specialist knowledge. Appropriate sourcing can be a complicated issue, and these are general rules. Deciding whether primary, secondary or tertiary sources are more suitable on any given occasion is a matter of common sense and good editorial judgement, and should be discussed on article talk pages.

The above definitions and examples are themselves reliably sourced. The discussion text is largely taken from the original with the flab or irrelevant material removed (for example, the use of reliable sources is a topic for another policy page). The short second-last paragraph is new, and contains material that has come up in discussions. We need to avoid absolutist categorisations of sources, as that leads to confusion. Colin°^Talk 23:23, 11 January 2009 (UTC)

There are several issues that make the proposed language problematic.

Journal articles are presently only considered primary sources for PSTS if they present experimental results. If they present original research in the form of analysis or synthesis they are not considered primary for the purposes of PSTS.
I don't know what is the difference between a "first hand" newspaper account and a "second hand" newspaper account.

— Carl (CBM · talk) 01:15, 12 January 2009 (UTC)

First hand: The reporter saw the car on fire with his own eyes.

Second hand: The reporter talked to someone who saw the car on fire. WhatamIdoing (talk) 02:16, 12 January 2009 (UTC)

I don't see that one of those is more or less reliable in general than the other; either both should be treated as primary, or both should be treated as secondary. — Carl (CBM · talk) 02:43, 12 January 2009 (UTC)

Carl, reliability has noting to do with this. Let's stay focused on Original Research. Blueboar (talk) 03:05, 12 January 2009 (UTC)

The main motivation I can see for discussing PSTS on the NOR page is as a proxy for reliability, so I keep it in mind.

But ignoring reliability, I stand by the point that a news story in which the reporter interviews an eyewitness is not "removed" in any significant way from the incident. Moreover, regardless whether the reporter describes her own experiences or quotes another witness, the story will still be read by an editor before publication, and anything that seems surprising will be fact-checked. — Carl (CBM · talk) 03:17, 12 January 2009 (UTC)

There is another caveat that needs to be mentioned if we are to rework the PSTS secton... subject context. In the study of history, at least, the same document (and not just material within a document) can be either primary or secondary depending on what the topic of discussion is.

My favorite hypothetical example of this is a book on the history of Saxon England, written by a prominent Victorian era historian. If we are writing an article on the history of the British Isles, this book is a secondary source. A lot of the material would probably be outdated, but it would still be a secondary source.

However, in an article with a more historiographical focus, (say: How different eras viewed Saxon England or the Victorian attitudes and worldview) that exact same book becomes a primary source (and subject to all the limitations that the policy now contains). Blueboar (talk) 03:05, 12 January 2009 (UTC)

I must ask that if folk want to present a different interpretation of PSTS (e.g. Carl's "If they present original research in the form of analysis or synthesis they are not considered primary for the purposes of PSTS") then they back this up with sources. All sources I have found regard all primary research papers as primary sources. If the analysis or synthesis is of the researcher's own unpublished data, this is a primary source for those conclusions -- they are "original thoughts". Even so-called "experimental results" are the consequence of interpretation (the thermometer assumes if the voltage changes the temperature has changed; the researcher assumes that if the patient's temperature is down then he is better). And I wholeheartedly agree with Blueboar that making "PSTS on the NOR page is as a proxy for reliability" is a misguided approach that leads one to come up with "that's not a primary source; its reliable" sort of nonsense. A Simpson's episode is a very reliable source for what characters the Simpson's episode contained; it is still a primary source. Lastly, the text does not say '"second hand" newspaper account'. Please be precise. The bulk of newspaper text these days isn't by journalists in the field. It is a rehashing of press releases or opinion articles. The story you read about wonder drug X is a secondary source. The discussion on the Middle East crisis by someone who hasn't left Kent, is a secondary source on the situation in the Middle East. Colin°^Talk 08:07, 12 January 2009 (UTC)

Please consult the lengthy archives for discussions about PSTS. The only summary I can make of previous discussions is that there are several different viewpoints on the distinction between primary and secondary sources, but no general agreement, and that the wording used in NOR is idiosyncratic. Start by reading these sections:

The present list of examples of primary sources on the actual policy page includes "published notes of laboratory and field experiments or observations; experimental results written by the person(s) who conducted the experiments;" but it does not include other journal articles. However, PSTS says that secondary sources rely on "facts and opinions on primary sources, often to make analytic, synthetic, interpretive, explanatory, or evaluative claims." Many journal papers do exactly that, and thus meet our standard for "secondary sources".

This situation is somewhat intentional, because the present policy is listing a particular subtype of primary sources - the type that should not be used to source analytic or synthetic claims. Also, the PSTS section is (or has been) used to determine which sources are valid for establishing notability, and many journal articles count as secondary sources for the purposes of WP:N. — Carl (CBM · talk) 13:00, 12 January 2009 (UTC)

This is abusing PSTS. What on earth does "journal articles count as secondary sources for the purposes of WP:N" mean? Such a sweeping statement makes no sense. If we are talking about whether a person is notable, then a biographical article of some famous scientist (often an obituary) in a scientific journal would be a secondary source, sure. But a single paper written by that person would not, and would certainly not establish notability. I can't see how a category of publishing can always be primary/secondary/whatever for any possible fact someone may care to draw from it.

There seems to be some confusion that if writing involves analysis, interpretation, evaluation, etc, that it cannot be primary. This is wrong. The key is whether the writing is about previously published work. The above cited discussions continue to perpetuate the flaw that WP editors are just stating their own personal definitions and refusing to agree. Let's stop that and have a source-based discussion for once. Please provide sources that disagree with the above proposed text. Colin°^Talk 13:45, 12 January 2009 (UTC)

The nutshell for WP:N says, " If a topic has received significant coverage in reliable secondary sources that are independent of the subject, it is presumed to satisfy the inclusion criteria for a stand-alone article." Many journal articles are included in the set of "secondary sources" that are used to establish notability. That's what I mean when I say journal articles count as secondary sources for the purposes of WP:N.

If you insist, I will point out that the proposal above designates a source as primary if it includes "original thinking that has not been previously published, ...". This could be read to say that a monograph of new scholarship about the French Revolution, which proposes a new interpretation or explanation of of the revolution, counts as a primary source. That seems to contradict the first and third sources that the proposal itself provides (the second source only claims to describe the health sciences). Moreover, both the first and third sources provided by the proposal include include "journal article" as primary and secondary sources: [5] [6]. The latter directly says "particularly in disciplines other than science" to count journal articles as secondary sources.

My analysis of this is that there are two dimensions being considered: "originality of thought" and "analysis and synthesis", rather than a single dimension "primary/secondary". — Carl (CBM · talk) 14:25, 12 January 2009 (UTC)

The WP:N guideline requires secondary sources for the situation where the primary source is the subject being scrutinised for notability. For example, is a book, play, painting notable? The issue of PSTS is pretty irrelevant wrt real world things such as diseases, chemical reactions, galaxies, etc. Concerning science, if someone created an article on a scientific study (and there are a few notable such studies) then the primary source documenting that study would be insufficient to determine notability. However, if the study is widely cited, then there are plenty secondary sources (also journal articles) that would suit that purpose.

In your French Revolution example, the monograph is a secondary source on the F. Revolution. It is a primary source on Professor's Smith's opinions.

Please be careful with saying "Journal articles are...". I've done no such thing. Some journal articles are primary sources, and some are secondary sources. Most humanities journal articles will be secondary sources, because they describe and analyse existing primary sources (literature, artefacts, etc). In scientific research, someone actually did a real-world experiment and discovered something new. And it doesn't matter how much analysis the person did to reach that novel conclusion, or how many log books they keep and revise. The first time they publish that research, it becomes a primary source for the research. Colin°^Talk 16:56, 12 January 2009 (UTC)

The proposed text says, "journal articles presenting original research", and seems to imply that the "original research" there includes not only experimental data, but also analysis or synthesis. So it appears that a monograph on the French revolution would be classified as secondary under the proposal, but the same monograph would count as primary if it were abridged and published as a paper in a journal. That seems extremely counterintuitive to me. — Carl (CBM · talk) 18:15, 12 January 2009 (UTC)

I've clarified those two examples. See below Colin°^Talk 18:46, 12 January 2009 (UTC)

I think we are (once again) faced with the problem of having different academic fields define the terms primary and secondary very differently. An original journal article is considered a primary source by those who work in the Sciences, and a secondary source by those who work in the Humanities.

The important thing to me isn't really whether a source is primary or secondary... it's whether we as Wikipedians are misusing the source in a way that introduces OR into an article. It is possible to misuse any "type" of source... but because it is easier to misuse a primary source, we need to include a special caution about primary sources and list a few simple examples to get the idea across.

Our job is not to cover every single possible way primary sources can be misused, nor to list every item that could be considered a primary source... our job is to convey a broad principle (Don't insert your own thinking into an article).

I have always had a problem with how the PSTS section ties in with the rest of the NOR policy... The rest of the policy focuses on how editors can misuse sources, and the PSTS section suddenly shifts that focus to a discussion of the type of sources. That transition is jarring... and tends to make people think that we are somehow banning the use of primary sources (which is not the intent). I think it would help if, in any rewrite of PSTS, we can figure out how to keep the focus on the misuse of sources. Blueboar (talk) 15:34, 12 January 2009 (UTC)

I don't think there's a conflict between fields. I don't think the term "journal article" is particularly useful on its own in this discussion. Much like "newspaper article" is insufficient in order to determine whether a source is primary or secondary. Colin°^Talk 16:56, 12 January 2009 (UTC)

Unlike "journal article presenting original research"? — Carl (CBM · talk) 18:15, 12 January 2009 (UTC)

OK. I've added "scientific" prefix to the "journal articles presenting original research". And added "humanities journal articles" as typically secondary sources. The precise wording of the examples can be tweaked, but should still be source-based rather than WP-editor-belief. The absolutely vital thing is that the definition of PSTS are sourced-based and reasonable. The examples flow from the definitions. Once you go the other way round, you are effectively conducting original research on your own definition. Colin°^Talk 18:46, 12 January 2009 (UTC)

The definitions in the current policy text are truly lame. We have "Primary: close to an event", "Secondary: one step removed from an event". This is at the Baldrick level of definitions ("Dog: not a cat"). How can someone apply those definitions to a novel, map or a 20-year scientific study. Plus we have the nonsense that "The key point about a primary source is that it offers an insider's guide to an event, a period of history, a work of art, a political decision, and so on." I'm afraid the "work of art" is the primary source and doesn't offer any insider's guide to itself. So far, nobody has offered a serious source-based attack on the definitions proposed above. Just personal opinion and distraction about reliability, proxies for this or that, or what peer-review means for PSTS. Colin°^Talk 18:46, 12 January 2009 (UTC)

The proposed text includes "original thinking that has not been previously published" as part of the definition of primary source, even though the websites it links to indicate that certain types of previously unpublished thinking are secondary sources. For example, biographies may include novel interpretations of a person's motivations, and monographs may present and argue in favor of novel theories, but these are not primary sources. The very websites cited in the proposed text contradict themselves on this point. So the problem is not that the proposed text is completely at odds with sources, it's that the sources themselves are not particularly clear, and cannot be if they want to be general enough to apply broadly to many fields. This has been discussed at great length in the archives, which as included long lists of websites on primary/secondary typing. — Carl (CBM · talk) 18:57, 12 January 2009 (UTC)

In so far as the biography or monograph contain facts that can be traced to the primary sources used by the authors, the biography and monograph are secondary sources for those facts. Once the author starts offering original opinions (on motivation, say) or original theories, then those opinions and theories are disconnected from the primary sources used, and the biography or monograph becomes a primary source for the opinions and theories. What you seek is a simple rule that says "source X is primary" or "source Y is secondary" for all text within the source, for all purposes and for all time. That rule doesn't exist. Colin°^Talk 19:37, 12 January 2009 (UTC)

If there is no firm line between primary and secondary sources, perhaps NOR should say that instead of attempting to define primary and secondary sources. But the proposed text does not say that. It's also not clear what is permitted when secondary sources are used that is not permitted when primary sources are used.

Back to the revolution example: If somebody writes a new monograph on the French Revolution, which includes original and previously unpublished analysis of the revolution, and we use it in our article on the French Revolution to present the theory that the author is presenting, this is a use of a secondary source. Do you agree? — Carl (CBM · talk) 19:46, 12 January 2009 (UTC)

I'm going with the "original: completely new and not copied or derived from something else"[7] definition. If the theory is grounded in the primary sources used by the monograph, it isn't truly original (merely new). Once the theory becomes speculation (suppose X.. then Y and Z make sense) then it may well be original but the monograph starts to become opinion and can be cited as a primary source of that opinion. I accept that in history the dividing line between opinion and fact can be blurry. Colin°^Talk 21:02, 12 January 2009 (UTC)

I'm sorry, I really want to stick with this, because I think it may either expose some common ground or expose a more fundamental disagreement. I am talking about the specific WP article on the French revolution, not an article about some particular theory. The topic of the WP article is the French revolution.

Suppose that there is a particular paper from the late 20th century that presented a previously-unpublished interpretation of some event of the French revolution. Suppose that I cite that paper for the part of the article that describes the interpretation presented by the monograph. I would say that I am using the monograph as a secondary source at that point.

I would argue further that it is completely impossible for any new primary sources on the subject of the French revolution to come into being. Any newly written source about the French revolution must be a secondary source, since the revolution ended over a hundred years ago. Compare: [8] [9]. The latter is particularly interesting to me:

"[Secondary sources] are usually in the form of published works such as journal articles or books, but may include radio or television documentaries, or conference proceedings."

— Carl (CBM · talk) 21:24, 12 January 2009 (UTC)

I'm not a humanities expert so you've got me out of my depth here. You've lost me with the paper/monograph bit. Which are you citing and which is based on which?

Looking briefly at the French Revolution article I see there's some recent speculation that an El Niño lead to a famine that sparked the whole thing off. If on WP you wrote "The French Revolution was caused by famine resulting from an El Niño" and cited a scientific research paper first proposing this original idea, then it would be a primary source for the El Niño aspects of your article. For the famine aspect of your article, it would be merely one of many possible secondary sources.

I'm not sure what is "interesting" about the quotation. How does the publication format (book, radio, TV, Web) affect whether it is primary or secondary? Colin°^Talk 21:59, 12 January 2009 (UTC)

I think that what I had hoped was an area of agreement may actually be an area of disagreement. The article by Groves that speculates on El Niño and the French Revolution is still a secondary source for the article on the French Revolution, because it's impossible to create a new primary source about the French Revolution. Perhaps of a different WP article, The El Niño theory of the French Revolution, Grove's research article would be a primary source, But for the actual article French Revolution the journal article by Grove is a perfectly relevant secondary source - it analyzes primary sources to provide an interpretation of the events, and it was greatly separated in time from those events. There may be other reasons, such as undue weight, not to emphasize Grove's theory, but "primary source" cannot be one of them.

The thing that is interesting to me about the quote (also in the sources listed below) is that journal articles are classified as secondary sources. That question has been discussed on this talk page quite a bit recently. I am supporting my claim that "Journal articles are presently only considered primary sources for PSTS if they present experimental results." by presenting sources, as you asked. — Carl (CBM · talk) 22:11, 12 January 2009 (UTC)

On Grove: this is the kind of dispute that doesn't have anything to do with OR (though the claim describing the research might be slightly exaggerated). The point of the PSTS in this policy isn't to categorize every source into whether it's primary, secondary, etc. The objective here in this policy is delineate and eradicate wikipedian original research. If the primary, secondary, tertiary source says, "El Niño caused the famine", then it's not original research to cite the claim here. If the primary source says, "El Niño changed worldwide weather patterns between 1789 and 1793", and the wikipedian says, "El Niño caused the famine", that would be original research. Professor marginalia (talk) 22:55, 12 January 2009 (UTC)

Very true. However, I think it safe to be able to define PSTS without the baggage of what people are going to do with the categorisation. This is why I despair when I see PSTS being altered or stuck with one form merely so it can prop up WP:N or WP:V, etc. BTW: "experimental results" is too restrictive. I'd prefer "scientific journal articles presenting original research". For example, an epidemiological study does not involve an experiment. Analysis of in-patient injury patterns does not require an experiment. Colin°^Talk 23:11, 12 January 2009 (UTC)

I don't think that we can separate the definitions in PSTS from the purpose we want to use them for. The definitions we pick are certain to depend on the use that we have in mind for them. The fact that PSTS is used in WP:N may be unfortunate but I think it's too ingrained to change at the moment.

Re "experimental", one distinction is between "lab science" and "theoretical science" that develops models axiomatically rather than gathering data. Theoretical papers often fit more closely into the "secondary source" mold, since their role is to analyze patterns form other research by providing frameworks for those patterns. — Carl (CBM · talk) 23:22, 12 January 2009 (UTC)

Actually, there isn't a problem with WP:N's request for secondary sources. This is obviously a good idea when the primary source is the topic being analysed for notability. But it is also true for other primary sources such as a diary of a person or a primary research paper. The latter doesn't establish notability of anything. It is only once other scientists start citing that research that the research (or researchers) become anyway notable. I can live with "scientific journal articles reporting experimental research results" if you insist since they are just examples, not an exhaustive list. Colin°^Talk 21:24, 13 January 2009 (UTC)

There are many topics that are only presented in research papers, but not in textbooks, and which would never be deleted in an AFD debate. Since WP:N requires "secondary sources" there are a few possible resolutions of this dilemma:

WP:N is worded poorly
Journal articles sometimes count as secondary sources for the purposes of WP:N

I find the second of these more pleasing. — Carl (CBM · talk) 13:07, 14 January 2009 (UTC)

You continue to say "journal articles" as though we've put them all in the primary source pot. We haven't. Please explain how a "scientific journal articles reporting experimental research results" can be used to establish the notability of any topic for which we will consider it a primary source (for example, the results, the study itself, and the journal article itself). Note that such a journal article is a secondary source for facts discussed in the "prior work" section, which cite earlier papers (as indicated by the proposed text: "Within one source document may be both primary and secondary material."). Colin°^Talk 18:01, 14 January 2009 (UTC)

Some quotes from online sources

Since it was asked whether merely consulting sources can clarify primary/secondary source typing, I looked up a selection of sites on primary/secondary sources that classify journal articles as secondary sources. Any emphasis here was added my me.

From U.C. Berkeley [10]: "A secondary source is a work that interprets or analyzes an historical event or phenomenon. It is generally at least one step removed from the event. Examples include scholarly or popular books and articles, reference books, and textbooks."
From Indiana U [11]: "History textbooks, dictionaries, encyclopedias, interpretive journal articles, and book reviews are all examples of secondary sources."
From Sonoma State U: [12]: "Secondary sources analyze, interpret or comment on the primary source materials. These include books, encyclopedia articles, critical essays, articles, reviews, dissertations and more."
From U. Michigan [13]: "In contrast, secondary sources are works that interpret or analyze the content of the primary sources. Most reference books, text books, and scholarly publications are secondary resources ... The key to determining whether a source is primary or secondary is the amount of time between the events recorded in the document and the time the document was created. "
From U. Victoria [14]. Describes "A scholarly article interpreting the symbolism of a kiss in Dante's body of literature" as a secondary source.
From U. Albany [15]: "The most important feature of secondary sources is that they offer an interpretation of information gathered from primary sources. Common examples of a secondary source are: ... Biographies ... Dissertations ... Indexes, Abstracts, Bibliographies (used to locate a secondary source) ... Journal Articles ... Monographs"
Eastern Connecticut U. [16] "Examples of secondary sources are biographies, journal articles, and documentaries written some time after the event being studied."

— Carl (CBM · talk) 21:45, 12 January 2009 (UTC)

Thank-you. I note most of those sites lean towards the humanities.

Your Indiana U site says "A column in the Op/Ed section of a newspaper ... is considered to be a secondary source... Bear in mind, however, that primary and secondary sources are not fixed categories. The use of evidence as a primary or secondary source hinges on the type of research you are conducting. If the researcher of the 2000 presidential election were interested in people’s perceptions of the political and legal electoral controversy, the Op/Ed columns will likely be good primary sources for surveying public opinion of these landmark events." This is the point made in the second-last paragraph in the proposed text above (which also has yet another citation to the same point). This point is also repeated in your U. Albany site. I think it is a valuable point (context is necessary) to add to the policy to help clarify things.
The U. Michigan site says "In English, History, Social Sciences, Literature and most other liberal arts, primary sources are works that were created during the time period about which they are written or by eye-witnesses of an event. In Science and Mathematics journal articles that discuss new discoveries, as opposed to summarizing previously written material, are considered primary sources, along with all the materials included in the liberal arts primary source list.
The U. Victoria site gives examples of primary sources as "Articles containing original research, data, or findings never before shared" and goes on to give the specific example of "A scientific article reporting on the growth rates of Douglas fir trees on Vancouver Island.".
From Timken science library [17]: "Primary sources present information that has not been previously published in any form in any other source. These sources may evolve through either formal or informal channels of communication. Journals, patents, and technical reports are examples of primary literature that have been evaluated through a peer-review process and are disseminated through published sources. Other primary sources, such as laboratory notebooks, memoranda, e-mail or listservs, are not usually published, but are nevertheless an important resource." [18]: "Secondary literature sources, including indexes and abstracts, encyclopedias, handbooks, reviews [Colin: which are often published in journals], and other reference sources, facilitate the assimilation of information originally disseminated through the primary literature. They integrate and improve access to the ever-increasing body of primary literature by organizing, repackaging, compiling and editing primary sources."
From James Cook U [19]: "Sources of information are generally categorised as primary, secondary or tertiary depending on their originality and their proximity to the source or origin.", "Primary sources present original thinking, report on discoveries, or share new information.", "They are usually the first formal appearance of results in the print or electronic literature", "They present information in its original form, neither interpreted nor condensed nor evaluated by other writers." Primary sources include "scientific journal articles reporting experimental research results". "A newspaper article is a primary source if it reports events, but a secondary source if it analyses and comments on those events." Secondary sources "are works which are one or more steps removed from the event or information they refer to, being written after the fact with the benefit of hindsight. " including "journal articles, particularly in disciplines other than science".
From Online Dictionary for Library and Information Science (ODLIS) [20]: Primary sources: "In scholarship, a document or record containing firsthand information or original data on a topic, used in preparing a derivative work. Primary sources include original manuscripts, periodical articles reporting original research or thought, diaries, memoirs, letters, journals, photographs, drawings, posters, film footage, sheet music, songs, interviews, government documents, public records, eyewitness accounts, newspaper clippings, etc." [21] Secondary source: "Any published or unpublished work that is one step removed from the original source, usually describing, summarizing, analyzing, evaluating, derived from, or based on primary source materials, for example, a review, critical analysis, second-person account, or biographical or historical study."
From U. Connecticut [22]: "The initial release of information in a formal setting is considered primary literature. Primary sources present information which has not been previously published in any form in any other source. In scientific literature, journals, conference proceedings, and technical reports are usually considered primary literature. Also, data upon which these publications are based can be considered a primary source.", "Secondary sources describe, comment on, interpret, analyze, summarize, or evaluate primary sources. They rely on evidence previously published in primary sources and are written after the time period when the experiment was conducted or the event noted in the primary source occured. ... Publication types that are usually secondary sources in the sciences include books, review articles, newspaper articles, and indexing/abstracting services"
From Washington State U. [23] "Primary Source: A primary scientific paper is defined as the first publication of original research results that is in a form with sufficient detail whereby the author’s peers can critically evaluate the research process and could repeat the study to test the conclusions.", "A secondary source is an information source that does not have as a major component the description of formal observations or experiments but rather is synthesized from some combination of primary sources, experience, or authoritative belief (dogma)."

I don't see any of those sites contradicting each other. I'll concede that journal articles in the humanities are generally placed in the "secondary source" bucket but assert that in the sciences there are two clear types: articles publishing original research are primary sources, articles based on previously published research are secondary sources. Plus there's further evidence of the point that even a secondary or tertiary source is also a primary source on itself, its author or its author's profession.

Can we find some way of tweaking the proposed definitions and/or example text so that it fits these sites, which are more authoritative than your opinion or mine. Colin°^Talk 22:53, 12 January 2009 (UTC)

After watching the discussion about primary/secondary source distinction for some time, I have the conclusion that there really isn't a solid definition that is generally applicable. But I also still don't know exactly why we are defining primary and secondary sources on the NOR page, and without that knowledge I don't know what sort of definition would be the most useful. Can we button that up first? — Carl (CBM · talk) 13:37, 13 January 2009 (UTC)

You know, what you've just said really implies you should recuse yourself from the discussion. Really the problem with the PSTS discussions is (a) people coming up with their own definitions, which confused everyone and (b) people saying the issue is confused when it really isn't and (c) people seeking a use for PSTS before they have established proper definitions. Defining a word one way just so that you can use it a certain way, is an Alice in Wonderland approach to language. The sources above (both yours and mine) are completely in agreement. The humanities and sciences classify their journal articles differently but this is easy to deal with and still broadly fit the definitions of primary vs secondary. What happens with this talk page is that ideas are discussed to death and we still have a policy page that is poorer than the proposed alternative. Compared to just about any of the sources cited above, our current policy page definitions of PSTS are rubbish and embarrassing. Colin°^Talk 21:24, 13 January 2009 (UTC)

Colin, I like most of your proposal, but I'm not keen on the "Within one source document may be both primary and secondary material" paragraph because, as I've argued before, I want to see better understanding that the policy is about "original research", and reduce the misapprehension that it's primarily a "don't use primary sources" rule. And I've never been in a dispute where I think elaborating over what kind of secondary sourced content can appear in a primary source document would have helped much. I can't remember one now at least. Professor marginalia (talk) 22:21, 13 January 2009 (UTC)

Well let's drop that and discuss it later. It is a point mentioned by at least three of the cited definitions and one I have seen people trip up on. Absolutism wrt to saying "X is a primary source" leads to folk getting all confused and thinking there is no solution. But it can wait. Colin°^Talk 23:55, 13 January 2009 (UTC)

New edit

Xasodfuih recently added the following to the PSTS secton: A scientific paper that presents original results often includes a section, typically the introduction, where related work is reviewed. This may constitute a suitable secondary source in the absence of dedicated review papers in that area.

I have reverted... It does not matter whether the introduction of a scientific paper is primary or secondary... the point is that it is a perfectly good source for us to use as long as we don't misuse it in a way that violates NOR. Furthermore, there is no need to list every single example of "this is OK... this isn't". FOCUS people... this policy isn't about the sources... it is about what WE say about the sources in our Wikipeida articles. Blueboar (talk) 20:14, 13 January 2009 (UTC)

See my comments above. If the policy is not about sources, or about trying to divide sources into primary/secondary.... then why is that section in the policy?

What this policy is "about" is that our articles should not make claims, either explicit or implicit, that cannot be sourced to reliable publications. It isn't about "what we say about sources in our Wikipedia articles". Our articles are not about sources, they are about topics, and so the policy discusses what sorts of claims we can make about those topics. — Carl (CBM · talk) 20:24, 13 January 2009 (UTC)

Thank you blueboar. I know I'm repeating myself, but the policy is not about labeling sources as primary, secondary, or tertiary. The policy is to not engage in original research. The hairsplitting about what sections or passages of a scientific research paper might be categorized as "secondary" is just pedantry. We already have guidelines written that talk about the use of sources in science and medical articles. These niggly caveats just aren't helpful here in the policy page! Professor marginalia (talk) 20:31, 13 January 2009 (UTC)

Carl: "It isn't about 'what we say about sources in our Wikipedia articles'." Many original research disputes are over "what we say the sources show in our wikipedia articles." Professor marginalia (talk) 20:38, 13 January 2009 (UTC)

As for why the PSTS section is in the policy in the first place... time for a history lesson:

The section started out as a simple one sentence statement: "Wikipedia should not become a primary source for information." It was thought that this statement needed further explanation, so people understood what the policy meant by "primary source." This in turn led to explaining what a "secondary" (and eventually "tertiary") source was. But... and this is important... the original intent was to make it clear that WIKIPEDIA should not become a primary source.

Unfortunately, as people started to add examples to cover their own areas of expertise the focus shifted. People began to debate the examples and whether they were or were not primary/secondary. We lost track of WHY we were discussing primary and secondary sources... to the point that, in one revision or another, someone took out the very sentence that caused us to have the discussion in the first place... the sentence that tied the discusion to the rest of the policy. The debate over definitions had taken on a life of its own.

We need to regain the original focus this section started with (and which the rest of the policy still has)... focus on the use and misuse of sources... not on whether a source is primary or secondary. Get the section back to focusing on the concept that "Wikipedia should not be a primary source for information." Blueboar (talk) 21:12, 13 January 2009 (UTC)

Ha, ha, ha. Editors reject using scientific papers but conduct votes to decide if Joe the Plumber is or isn't a plumber (like that's not original research). I found it looking through the (otherwise fine) contributions of Kenosis. I'm not going to waste anymore of my time arguing policy; got articles to write. Xasodfuih (talk) 21:11, 13 January 2009 (UTC)

Who said anything about rejecting the use of scientific papers? Are we having the same conversation? Blueboar (talk) 21:14, 13 January 2009 (UTC)

What you say, Blueboar, doesn't seem to at all explain why there is a faction who adamantly insist that PSTS remain in policy. This dispute goes far beyond simple clarification of what "primary" means. (Nor does the bulk of what appears seem to actually clarify, given that the starting point is that "primary" means different things in different fields. OTOH, "Don't introduce your own new ideas" seems pretty clear, as does the concept that Wikipedia is, by design and intention, not the place for any new idea to appear.)Minasbeede (talk) 21:28, 13 January 2009 (UTC)

I'm not sure how that phrase is helpful. You'll get people claiming that because they are citing primary sources (say) that they build their ideas upon that that makes the WP text a secondary source for the information.

I know you guys are weary of the PSTS thing but I'm trying, in the above proposed change, to base the section on external definitions. I think this has been achieved and contrary to expectations, there's no great clashing of humanities vs sciences that shall never be resolved. Once we have these definitions placed on a solid sourced foundation, future attempts to alter or re-debate the text can be closed off with "show us some reliable sources that say otherwise". Just like with our best articles -- they become stable when they are well cited. The current section is vastly deteriorated compared to what it used to be. It's backwards for goodness sake (yes I'm aware of the discussion that changed that). It needs to be improved and based on reliable external definitions. It should not be abused with baggage from other policy or guideline requirements. You will never get peace with the current section text -- it is just too awful and inadequate. Colin°^Talk 21:51, 13 January 2009 (UTC)

Sorta weary. There is WP:RS, which is the policy on reliable sources. Properly, PSTS belongs there if it belongs anywhere. Unfortunately there are some who regard some uses of primary sources as OR, so they insist on PSTS and insist it be in WP:NOR. For a very long time there has been discussion of PSTS in this talk page and resolution seems no closer now than it ever has. Minasbeede (talk) 23:18, 13 January 2009 (UTC)

I would certainly be one of those who feel that some uses of primary sources is OR... what is unfortunate is that there are some people who seem to feel that any use of primary sources is OR... and that is not what the policy says (nor what the policy should say).

We will get no where if we attempt to remove PSTS. That is an argument that has been done to death. There are people who feel that it is vital to the policy (as opposed to my view, which is that it is useful, but not vital) and who will ardently oppose any attempt to remove or demote it. So... instead of removing or demoting it, I am proposing that we keep it, but try to refocus it back to where it was originally. Blueboar (talk) 05:59, 14 January 2009 (UTC)

The examples that I have seen presented of how primary sources can be used in a way that violates NOR don't actually involve the primaryness of the sources. The examples involve editors who add conclusions to the article that are not supported by the literature; bu these examples of NOR violations would continue to hold if the sources used were secondary instead of primary. Maybe the issue is that I still haven't seen an example where it was literally the primaryness of the source that made a certain claim OR (but not an issue of undue weight). Phil Sandifer's post of 00:22, 14 January 2009 (UTC), below, explains this in more detail. — Carl (CBM · talk) 12:42, 14 January 2009 (UTC)

"I would certainly be one of those who feel that some uses of primary sources is OR." Sure. I agree. (Any type of source can be so misused.) The proper issues here are whether PSTS does anything to reduce the such occurrences and whether PSTS has unneeded and unfortunate side effects on material that is not OR. Does it create confusion for editors? (The answer to the last question is, of course, "Yes.") It is difficult to see how a policy section that is still misunderstood makes a positive contribution to Wikipedia. PSTS is so clumsily worded that it is hard to imagine it gives useful guidance to any editor. All that remains is for PSTS to be used as grounds for another editor to remove material - which happens, and happens without any explanation on the talk page. Let me be fully in accord with the principle of good will and assume and assert that all such material removed deserved to be removed: the removals are righteous. Was PSTS in WP:NOR needed to make such removal possible? If the removed material constituted OR was it necessary to have PSTS in WP:NOR to make the removal possible?

WP:NOR isn't just an article, isn't a guideline, it's a policy. The unfortunate thing is that PSTS makes it a policy that is unclear and which, no matter how worded, is jarring to many editors. The "people who feel that it is vital to the policy" surely are free to believe that yet do they have some obligation to make PSTS clear enough in its wording that the discussions no longer occur? You have put a lot of your time, thought, and effort into this, Blueboar, yet the problem persists. Might it be that there is actually no stable way to have PSTS be part of WP:NOR? Minasbeede (talk) 00:41, 15 January 2009 (UTC)

actual source for this

Depending on the quality of the journal, such reviews may constitute a usable source. Certainly the National Library of Medicine , Web of Science, and many other indexes code long articles with extensive references a "review", along with articles specifically labeled review in the journal, or from journals devoted primarily or entirely to reviews. Obviously, there is always a possibility of such a review being partial or biased or selective or unreliable, but this is also true of all other review articles. So i have always taught my students as a biomedical librarian, and teacher of librarianship, at Princeton and Rutgers, just as was taught to me at Berkeley. And so the textbooks have it also, in all sciences--from literally the first one I picked up, the standard book for chemists, Maizell, How To Find Chemical Information (2nd ed), p.205 has a perfect statement:

"Not all reviews appear in journals or books specifically devoted to the subject. A good review can appear in almost any chemical journal or other information source. Reviews are frequently labeled as such... however, any article or book with an extensive list of references, and a good discussion of these, is potentially a review,, whether or not it is so labeled. Patent sources can also provide an excellent review of prior art." DGG (talk) 21:52, 13 January 2009 (UTC)

I continue to not see the sense of this. It looks like the issue is one of undue weight, of failure to attribute tentative results to the people offering them, and to properly frame things. It's not a matter of what sources should and shouldn't be used. The current language - including the change to include "experimental results written by the person(s) who conducted the experiments" - is a ridiculously broad overreach that renders the whole of scientific literature primary source material, and thus heavily restricts its use.

This should be rolled back to [24]. The expanding of PSTS to cover seemingly all scientific journal articles is ridiculous. Phil Sandifer (talk) 22:06, 13 January 2009 (UTC)

"Actual evidence" of what? Sorry, I'm confused. Professor marginalia (talk) 22:23, 13 January 2009 (UTC)

sorry, I reworded the heading, I meant an actual published citable academic source (Wiley) that such articles do count as reliable sources. and if you examine most journal articles, they have a pretty minimal introduction. Phil, you may not see the sense in it, but it nonetheless is the accepted standard in all academic fields. I have , btw, just written to the PubMed people to verify their exact current indexing standards for using the designation "review" DGG (talk) 05:07, 14 January 2009 (UTC)

Experimental results language

I continue to not see the sense of this. It looks like the issue is one of undue weight, of failure to attribute tentative results to the people offering them, and to properly frame things. It's not a matter of what sources should and shouldn't be used. The current language - including the change to include "experimental results written by the person(s) who conducted the experiments" - is a ridiculously broad overreach that renders the whole of scientific literature primary source material, and thus heavily restricts its use.

This should be rolled back to [25]. The expanding of PSTS to cover seemingly all scientific journal articles is ridiculous.

Can someone point me to the talk page section where this change was agreed upon? Phil Sandifer (talk) 22:06, 13 January 2009 (UTC)

The main problem with this is that the following paragraph says, "Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is verifiable by a reasonable, educated person without specialist knowledge". This is completely counterproductive, especially because of the value of journal results in scientific topics. This was a bad change, IMHO. Bastique ^demandez 23:26, 13 January 2009 (UTC)

"It's not a matter of what sources should and shouldn't be used"-it doesn't say this either. We're going round and round. Professor marginalia (talk) 23:46, 13 January 2009 (UTC)

Yes it does. Because it says, very clearly, in PSTS, that "Wikipedia articles should rely mainly on published reliable secondary sources and, to a lesser extent, on tertiary sources." So any time something gets labeled as a primary source, we are explicitly saying it should be used more sparingly. Phil Sandifer (talk) 00:05, 14 January 2009 (UTC)

First: "scientific journal articles reporting experimental research results" are considered primary sources by all authoritative definitions I have found. That you may be uncomfortable with the consequences of this is a secondary issue. We must stop abusing these words to mean something that suits an agenda. We define them the way authoritative sources define them. There's really no option. And extrapolating "scientific journal articles reporting experimental research results" to "seemingly all scientific journal articles" is itself ridiculous hyperbole. If the following paragraph (about descriptive claims) breaks, in some peoples minds, due to the change then perhaps it is that paragraph that is wrong? Would you guys listen to yourselves. You cannot define PSTS to fit current sentences in this or any other policy. The Earth is round. Deal with it.

Second: many expert writers on WP believe "scientific journal articles reporting experimental research results" should be cited sparingly. And I'm talking about mainstream editors who believe in science. There are good NOR reasons for preferring secondary sources when describing the "facts" learned from a research study. But even supposing your view had consensus (which, from the discussions we've had at WP:MEDRS, it doesn't) then the wording of any restriction to place on such primary sources should be examined. Not the definition of PSTS itself. Colin°^Talk 00:15, 14 January 2009 (UTC)

My problem is that I don't see why NOR needs to be broadened to get the result you want. If what you want to stop is overbroad conclusions from journal articles, first of all, WP:V already lets you go "Source doesn't say that" and take it out. After that hurdle you have "This gives undue weight to a viewpoint espoused by one paper." And after that, you have the fact that a claim along those lines should be narrowed and sourced - not "X works in treating Y," but "In Paper Z, X was found to have the following effects with regards to Y, but further research in the area (has not been done/did not verify the results/whatever is applicable)." Remembering, at each step, that the onus is on the person who wants to add the information to come up with a workable formulation of it. How on Earth, given all of that, do you need a modification to NOR to solve this problem?

I'm glad you have consensus on MEDRS, but that is a subject area content guideline. It does not spill out into other sciences, little yet to a site-wide policy. Phil Sandifer (talk) 00:22, 14 January 2009 (UTC)

I think it would be good to stop using science and medicine then to predict calamitous consequences of the PSTS clause in NOR. This policy says "don't make novel claims from primary sources." It emphasizes it. Why? Because in academics researchers, writers, analysts, etc are encouraged to rely on primary sources to come up with some new insight, analysis, etc. This is what secondary sources normally do. But this is not what wikipedia is allowed to do. And a lot of editors don't get it, wouldn't get it, unless a policy told them. This policy (tries to) explain this-and make it plain, this Is Not The Place to share your own insights about what it says in the Bible, or in this month's latest research finding in autism, or your discovery that the lyrics in Stairway to Heaven are the same as the words in Leaves of Grass in reverse order. These are clearly examples of original research from primary sources. And continue to occur every day here at wikipedia. Professor marginalia (talk) 00:51, 14 January 2009 (UTC)

I agree, but on the other hand it's idiotic to suggest that there is anything particularly tricky about summarizing what it says in an academic source. We shouldn't make novel claims from any sources. The problem is that people have some very puzzling ideas of what a novel claim is. Phil Sandifer (talk) 01:00, 14 January 2009 (UTC)

Phil, you say I am trying to change PSTS "to solve a problem". That's exactly what I'm trying not to do. I do have opinions on the use of sources, which have come up, but I'm really trying to get us to accept a definition of PSTS that is independent of our views on how sources should used and any caution we should apply. As long as this policy page disagrees with what people are taught at Uni about source classification, we will keep confusing people.

I think there is an issue with our rules concerning primary sources that needs to be clarified. We use primary sources in two ways: as sources about themselves, and as sources about something else. The former is the classic source=novel in an article about the novel. We have rigid rules about "descriptive claims" and require secondary sources to say anything more than that. This is good. But where the primary source is about something else (a diary used to source a claim about a person, a research paper describing a potential new cause of autism used to source a claim that the cause of autism has been found, an editorial used to source a comment about a newspaper's stance on an issue) then our NOR (and other policy guidelines) tend to be more advisory rather than preventative. This is not perhaps clear in our current wording and is possibly leading to Phil's problems with the inclusion of certain sources in the primary box. Colin°^Talk 10:03, 14 January 2009 (UTC)

"As long as this policy page disagrees with what people are taught at Uni about source classification, we will keep confusing people." The difficulty is that different people learn different things, depending on their field; there isn't any single "true" definition of primary and secondary sources.

For example, some people believe that if a newspaper writer interviews someone and then writes a news story, that makes the news story a secondary source. Others will believe that all news stories like that are primary sources. Some learn that a source that presents interpretation or synthesis is a secondary source by nature, while others are more concerned with the originality of the work. In some fields, like medicine, the distinction is more experiment/review than primary/secondary. In mathematics, the primary/secondary distinction is of little value, and we never use those terms. — Carl (CBM · talk) 13:00, 14 January 2009 (UTC)

Show me an authoritative definition that would allow the first interpretation of the news story being secondary. That sounds like someone is confusing second-hand with secondary source. If PSTS is relatively unimportant in the field of maths then fine. Let's have some evidence that "there isn't any single "true" definition" rather than just repeating it like a mantra. Colin°^Talk 13:27, 14 January 2009 (UTC)

I don't agree that there are "authoritative" sources here, but some examples are [26] and [27]. I disagree with the analysis of these, but I have seen it around. I think someone called this the "librarian" definition, vis-a-vis the "historian" definition. — Carl (CBM · talk) 17:00, 14 January 2009 (UTC)

"This policy (tries to) explain this-and make it plain, this Is Not The Place to share your own insights about what it says in the Bible, or in this month's latest research finding in autism, or your discovery that the lyrics in Stairway to Heaven are the same as the words in Leaves of Grass in reverse order." But this has nothing to do with the primaryness of these sources - you also should not share your own insights about a review paper, or your own insights about material covered in textbooks, unless those insights can be sourced. — Carl (CBM · talk) 13:03, 14 January 2009 (UTC)

You are right, but the reason we ask people to be much more careful with primary sources is that they contain fewer insights and often lack context (or people forget that context is important). So it is much easier to commit OR (or fill the gaps, or extrapolate the facts) when using primary sources than when using secondary sources. The same abuse is possible with either, but less likely with one. This is why the rules we have on primary vs secondary should often be advisory rather than fundamentalist. Colin°^Talk 13:27, 14 January 2009 (UTC)

It's not universally more likely to see OR from primary sources. In mathematics articles, the OR that I see is much more often from someone who looks through a couple textbooks and then tries to synthesize a new theory from them. It's very uncommon on WP to see someone look at numerous mathematics journal papers and do the same thing (not only because the journal papers are much harder for an untrained person to read). — Carl (CBM · talk) 14:13, 14 January 2009 (UTC)

I would also dispute the universal validity of the claim that secondary sources include more insights than primary ones. In other fields such as mathematics, new insights are primarily published in journal articles, while textbooks contain very few new insights compared to the research literature. — Carl (CBM · talk) 14:21, 14 January 2009 (UTC)

Can you give an exception other than maths? It seems that pure maths is an exception because ideas aren't based on texts or people... there are no tangibles being discussed or analysed. Colin°^Talk 16:29, 14 January 2009 (UTC)

There's also the humanities, and theoretical physics and theoretical computer science. — Carl (CBM · talk) 16:53, 14 January 2009 (UTC)

At least in literary theory, the textbooks are bloody useless. And in a fair chunk of hard sciences there's a "lying to children" problem with the textbooks, where simplified models of phenomena are given in intro books. Phil Sandifer (talk) 16:56, 14 January 2009 (UTC)

I confess this looks like a storm in a tea-cup to me. Both wordings make it clear that raw data from experiments, etc. is off-limits. From what I've seen of the people involved in this discussion, we'd all agree that conclusions based on these data are admissible if published in a reputable journal - although, if there has been insufficient time for the work to be replicated, scrutinised etc. by other scientists, it would be prudent to write e.g. "In YYYY author A concluded that P is the case" rather than simply "P is the case". All this is just common sense. However we probably have to clarify the whole WP:PSTS thing because it seems to generate more heat than light. --17:32, 14 January 2009 (UTC)

Remove PSTS altogether?

If we removed the whole PSTS section and replaced it with the following, how would that affect the way this policy is applied?

"Articles may include analytic, synthetic, interpretive, explanatory, or evaluative claims about published material so long as they have been published by a reliable source. For example, an article about a novel may cite passages from the novel to describe the plot, but any interpretation of those passages needs another source."

--Phenylalanine (talk) 23:49, 13 January 2009 (UTC)

My point is that discussion of PSTS really belongs at WP:RS IMO. We don't need to discuss terminology at all in this policy to explain the intent of the policy: "Articles may include analytic, synthetic, interpretive, explanatory, or evaluative claims about published material so long as they have been published by a reliable source." If editors want a detailed account of what kind of sources are most reliable and preferred, etc., they can look at Wikipedia:Reliable source. Maybe I'm wrong, but I don't see it. --Phenylalanine (talk) 12:15, 14 January 2009 (UTC)

At first I was sympathetic to Phenylalanine's "discussion of PSTS really belongs at WP:RS" (12:15, 14 January 2009). however when I checked I saw that WP:RS and WP:V don't really cover the primary / secondary / tertiary sources issue.

However it might be a good idea to transfer WP:PSTS to WP:RS (including the redirect), so that everything about selection and use of sources is in one place. My reasoning is that WP:NOR and WP:V are pretty easy to grasp in principle, but the detailed rules about use of suorcesd are not, and a one-stop-shop would be easier to look-up and to keep consistent. I'll cross-post to WT:RS and ask for discussion here (WT:NOR). --Philcha (talk) 12:33, 14 January 2009 (UTC)

Since my general belief is that the PSTS section here is intended as a proxy for reliability anyway, I have no objection to moving the PSTS section to WP:RS. — Carl (CBM · talk) 12:46, 14 January 2009 (UTC)

Can you explain this belief. What makes primary sources unreliable compared with secondary? Colin°^Talk 13:33, 14 January 2009 (UTC)

Higher on this page, you said, "... the reason we ask people to be much more careful with primary sources is that they contain fewer insights and often lack context (or people forget that context is important)." These are exactly the sorts of concerns that I believe motivate the PSTS sections – and these are all concerns about the reliability of the source for use on wikipedia, rather than questions about originality of research.

The NOR policy is that we cannot add any new insights, even implicitly, to our articles (regardless whether those insights were inspired by primary or secondary sources). So when we are comparing the use of primary sources vs. secondary sources, we have to be talking only about things that can be sourced, since otherwise the distinction is irrelevant. So, given that we are now talking only about things that are actually sourced, what reason is there to favor secondary sources over primary sources? The only reasons I can think of relate to reliability. — Carl (CBM · talk) 14:19, 14 January 2009 (UTC)

How is "fewer insights and often lack context" a reliability issue? Reliability is about whether a type of source is typically right or wrong. Please can you give some examples of where a primary source is typically more reliable than a secondary source? Colin°^Talk 16:32, 14 January 2009 (UTC)

Hang on folks, this discussion is supposed to about where where WP:PSTS should live, not about what the details of its content should be. --Philcha (talk) 16:45, 14 January 2009 (UTC)

The PSTS acronym pipes to the pertinent clause in OR. But distinctions about the benefit/use of primary, secondary and tertiary sources are found in other policy/guideline pages besides this one. Professor marginalia (talk) 16:50, 14 January 2009 (UTC) Examples, WP:Reliable sources#Primary, secondary, and tertiary sources, WP:Notability (fiction)#Secondary sources, WP:Notability (web), WP:WELLKNOWN, WP:Fringe theories#Identifying fringe theories, WP:Manual of Style (writing about fiction)#Primary and secondary information, WP:MEDRS. Professor marginalia (talk) 17:13, 14 January 2009 (UTC)

Goodness. That WP:Reliable sources#Primary, secondary, and tertiary sources section has got itself well confused. It says "Primary sources can be reliable in some situations, but not in others... Primary sources are considered reliable for basic statements of fact as to what is contained within the primary source itself (for example, a work of fiction is not a reliable source for an analysis of the characters in the work of fiction)." Reliability doesn't come into it. It just isn't a source for the "analysis" at all. How far has this "PSTS is a proxy for reliability" confusion gone? Colin°^Talk 17:26, 14 January 2009 (UTC)

That bit in WP:RS started as something I added... If you look through the discussions at RS and RSN, a lot of people were under the impression that primary sources were not reliable under any circumstances (and pointing to PSTS to back the argument, even though PSTS says nothing of the sort). I wanted to correct that by adding a line that essentially said that reliability and source type are not connected. Some primary sources are quite reliable, others are not. Unfortunately, there has been some instruction creap since I added my initial comment. Blueboar (talk) 18:11, 14 January 2009 (UTC)

I think you're right, Colin, that coming to at least a common, general, and widely accepted definition of primary, secondary sources will eliminate one area of confusion. With that, the policies and guidelines can focus on the wording of the advice in their use as relates specifically to that policy or guideline. Professor marginalia (talk) 18:30, 14 January 2009 (UTC)

Experimental results removal

Today the following was removed as examples of primary sources, "published notes of laboratory and field experiments or observations; experimental results written by the person(s) who conducted the experiments"[28], and described as a "controversial" language change in the edit summary. Can I ask, what evidence is there of any controversy? The language was there at least 15 months, until it disappeared in a general reshuffling of the order in which PS, SS, and TS were described. I can't find any controversy raised about it, at all, until in the last couple of days. And that concern was sorted out, from the looks of it to me. Published experiments are primary sources. There's no "controversy" over this. Guidelines for articles related to published experiments (science, medicine, fringe theories) do urge extreme care in their use, and aren't used as a basis for a novel claim. What controversy persists over this? Professor marginalia (talk) 18:17, 14 January 2009 (UTC)

Actually "written or recorded notes of laboratory and field experiments or observations" is the older language that had been in NOR for a while; maybe the removal of the second sentence was accidental? I agree that the "published experimental results by the person(s) actually involved in the research" is not very controversial, since it only refers to experimental results. — Carl (CBM · talk) 18:26, 14 January 2009 (UTC)

Random Qtrly versions: "published experimental results by the person(s) actually involved in the research" was there in September 08,[29] June 08,[30] March 08,[31], Dec 07,[32] and Oct 07[33]. Not there in Sept 07[34]. Professor marginalia (talk) 18:42, 14 January 2009 (UTC)

I have no problem with adding it back in. I have other concerns about PSTS in general, but not with that specific example. Blueboar (talk) 18:44, 14 January 2009 (UTC)

15 months? I reverted to a version from a few days ago, based on the discussion a few sections up in which multiple people expressed reservations about the "experimental results published by the person(s) actually involved in the research" language as seeming to broadly cover a huge swath of journal articles. The language was not in the page a month ago: [35] and, so I have trouble painting it as a long-standing revision. If someone wants to go through and find an explanation for its vanishing and return, I'd be interested in understanding what's going on here, but this seems to me language that was, at the very least, gone for quite a while, and language that multiple users have objected to.

Certainly I see no discussion establishing consensus for it. I'm fine with language about data and lab notebooks, but as phrased "experimental results" seems to me to render a huge swath of journal articles to be primary sources, and thus restricts their use to only what non-specialists can grasp. I am very much skeptical of the wider consensus for this. Phil Sandifer (talk) 19:03, 14 January 2009 (UTC)

I presented the links showing it's been there for 15 months. The first mention of it on the talk page was here, asking in effect "where did the 'experimental results' example that used to be here go?" (Wikipedia talk:No original research#Question about definition of a primary source.) It disappeared in this major restructure[36] (since essentially undone). A review of discussions on the talk page during that time frame don't show anyone objecting to it in particular. The editor who made the change said that all he intended was to reverse the order, and "I just want to remind people that my proposal is just to reverse the order. If anyone agrees with that but doesn't agree with the wording I have here - well, just change the wording!" Wikipedia_talk:No original research/Archive 39#Reverse the order of sources. Nobody complained at the time that "experimental results" needed to be removed. The words were just overlooked or ignored as unimportant. Professor marginalia (talk) 19:21, 14 January 2009 (UTC)

Well. Now we seem to not be overlooking them, and they seem rather unfortunately broad. Phil Sandifer (talk) 19:46, 14 January 2009 (UTC)

We who? And what problems, exactly, do they have with it? I'm asking for some focus in order to settle it and move on. Professor marginalia (talk) 19:57, 14 January 2009 (UTC)

Both Bastique and I raised objections several sections above. The objection is that, as phrased, it makes a huge swath of publications primary sources, and thus restricts them under the "no specialist knowledge" clause, which is very burdensome for such sources. Phil Sandifer (talk) 20:00, 14 January 2009 (UTC)

Could you please elaborate on your objections? I'm unable to find what "'no specialist knowledge' clause" refers to. --Ronz (talk) 20:21, 14 January 2009 (UTC)

To PS. As was pointed out, the clause isn't guilty of "making it sound like they're primary sources", the fact is "they are primary sources", in the academic definition. [37], [38], [39], etcetera. So the argument would be that an exception should be made in the category of "experimental results" to allow claims be made from them that are not verifiable except to those specialist knowledge? So let's be clear here. The published experiment is a primary source, whether it's listed as a specific example in the policy or not. Unless we're welcoming the misapprehension that it isn't a primary source, then why insist on leaving it out? Isn't it simply still the case that your real problem is "the specialist knowledge" clause? Correct? Professor marginalia (talk) 20:38, 14 January 2009 (UTC)

To Ronz. The wording is, "Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is verifiable by a reasonable, educated person without specialist knowledge." I think the objection raised was that if "published experimental results" are listed as examples of primary sources, then the "without specialist knowledge" rule applies to those as well. I think the objection is for a special exception of sorts for those sources typically written at a technical level for experts in science and medicine, which are the main two areas "experiments" are performed. Professor marginalia (talk) 20:49, 14 January 2009 (UTC)

Thanks! Yes, experimental results are primary sources. Should we have an exception for them? I'm not convinced. The review processes of the journals that publish such results vary tremendously, and journals are created for the specific purpose of allowing researchers to get around more strict reviews. --Ronz (talk) 21:33, 14 January 2009 (UTC)

Agree with Professor marginalia. Let's define PSTS properly and then deal with the consequential rules built round it. If they need adjustment, so be it. Colin°^Talk 21:47, 14 January 2009 (UTC)

Please can folk compare the current text with that proposed in #PSTS revised above. IMO both the definitions and the examples are better, and the rules aren't significantly changed other than to remove redundancy or repetition. Professor marginalia would prefer to drop the "Within one source document may be both primary and secondary material" paragraph as being too detailed/fussy. A number of third-party definitions have been cited above (by Carl and me) that are all broadly in agreement despite claims that no two people share a definition of PSTS. Even Carl's blurry example concerning newspaper accounts being possibly primary or secondary if based on unpublished interviews (ie. second hand), depending on who you ask, is not a big problem since the example given "first-hand newspaper accounts" is clearly primary in everybody's definition. Carl would prefer "scientific journal articles reporting experimental research results" to the "scientific journal articles presenting original research" of the proposal, and I can live with that.

So I ask again that we consider replacing the current woeful text. Not only would our current "Primary sources are sources very close to an event" definition get zero marks in any exam, it isn't even grammatical (comparing an object with a period of time). Tweaks and sourced objections are welcome but let's please ignore any "we can't make that a primary source, because that means this rule now says X" objections. We can discuss the rules once we've got the definitions right. Colin°^Talk 21:47, 14 January 2009 (UTC)

I think the proposed text is a definite improvement.

Because PSTS applies to just about every policy and guideline where sourcing is discussed (especially WP:V, WP:OR, WP:RS, and WP:NPOV), I'm wondering if it might be better to have a separate PSTS article, which can then be linked to and summarized in proper context in those policies/guidelines. Having PSTS as its own article would hopefully make it easier to define the terms while centralizing the definition, and allow us to discuss rules at length. See Wikipedia_talk:Reliable_sources#Transfer_WP:PSTS_to_WP:RS.3F for a related discussion.

Colin's edit summary also reminded me that I've always been surprised that WP:IS hasn't been expanded and promoted to the point where it's more helpful. Maybe it's an argument against making PSTS it's own article? --Ronz (talk) 22:54, 14 January 2009 (UTC)

I do not think the "revised" language is any better. Subjecting the bulk of science journal articles to only non-expert summary remains a ludicrous move. Phil Sandifer (talk) 22:59, 14 January 2009 (UTC)

Can you separate the two issues, what a primary source is from nonexpert-level verification is needed? There seems to be some consensus that lingering confusions over our description of primary source is confusingly complicating discussions about what is desirable in policy. I can't, for example, envision why we'd be interested in requiring "non-expert summary" for the verification in every area except primary sourced experimental results, so I think this is an illustration that conflating the two so much might lead to an even more incoherent policy. I'll try to explain it again-you've removed the example, but you didn't change the policy at all. The "non-expert summary" clause STILL applies to published "experimental results" because published experimental results are still primary sources. Professor marginalia (talk) 23:22, 14 January 2009 (UTC)

Well, for me, the issue is this - as long as the non-expert summary clause applies to primary sources, I think it is inappropriate to label, effectively, all science research as a primary source for our purposes. Regardless of a normal definition of primary source (which we have already departed pretty far from), it's just bad policy to restrict science research that way, and it gives way, way too much power to various science cranks - just think of what would happen on Creationism topics if Creationist editors could remove all the complicated science because it's journal articles that are being cited in ways their poor non-expert brains can't handle.

We need to be pragmatic, and this expansion of the policy is reckless. If we want to drop the non-expert summary clause, I'm fine with the expansion to cover experimental results. But the two cannot both be in there. Phil Sandifer (talk) 06:05, 15 January 2009 (UTC)

Phil, I think you need to study some of our best controversial articles and examine whether the are built on primary (by everyone's definition but yours) or secondary sources. Take Evolution and Autism. Click on the PMID links, for example, and PubMed will inform you that the vast majority are review articles in scientific journals. The mind boggles, in fact, at the thought of trying to build Evolution from primary research papers. This "can't use almost all science" nonsense has to stop. In areas such as Creationism, our best defence is the use if high quality secondary sources. I'm the last person to try to prevent us citing good science. Colin°^Talk 07:34, 15 January 2009 (UTC)

Phil, two points:

Edit warring over the "experimental results" example doesn't change the policy. It doesn't have any impact. Claims made about primary sourced experiments still must be verifiable to educated persons without specialized knowledge. I can't add my own translation of the Rosetta stone to an article even though there is no list of do's or dont's which list "Rosetta stone" specifically.
I've contributed to creationism and science articles. You have diagnosed the problem there all wrong. Creationism claims will never be sourceable to "experimental results" published in prestigious scientific journals. And creationists routinely mine for evidence against evolution, the Big Bang, radiocarbon dating, etc., through published primary research in the sciences. This is standard operating procedure in creationism. This is what creationists at the ICR and so forth were almost exclusively concentrating on for decades, if not still doing so. While science rarely, as in "almost never", conducts experiments about creationist theories. This policy, if nothing else, clamps down on the same kind of idiosyncratic interpretations of published research taking hold at WP. Professor marginalia (talk) 15:02, 15 January 2009 (UTC)

I'm sorry, are you now saying that translation from a foreign language is OR too? Phil Sandifer (talk) 15:36, 15 January 2009 (UTC)

An obscure language that no one has spoken in 2000 years? Yes. That would be OR, in my definition. Or do you think the "no original research" rule really isn't intended for cases like that of an anonymous wikipedian adding their own translations of otherwise unpublished decipherments of ancient hieroglyphic texts? Professor marginalia (talk) 17:32, 15 January 2009 (UTC)

I'm thinking, if nothing else, that the headaches associated with opening a translation as OR can of worms are utter madness. If the standard is that no one has spoken it in a long time, presumably Latin would also be out? Phil Sandifer (talk) 19:25, 15 January 2009 (UTC)

I think that the "slippery slope" absurdity could be applied to every policy in the wikipedia. NPOV opens the "can of worms" that we have to include every view ever published anywhere. RS opens the "can of worms" that the medical textbooks written in 1628 are more acceptable than published experiments in JAMA. IAR opens the "can of worms" that we don't need to follow any of the rules. We don't allow unpublished translations of 2000 year old hieroglyphics - we do allow editors to source texts in Spanish and French. WP hasn't withered and caved - maybe because most of its editors don't go to ridiculous extremes with policy, and if they do, they won't get away with it. Professor marginalia (talk) 19:51, 15 January 2009 (UTC)

A new approach to the PSTS question

At WT:RS several people have been discussing how the current PSTS section may be trying to do too much for too many different people... that source typing relates to several of our policies and guidelines... that the issue of appropriately using (or inappropreately misusing) different source types may need to be discussed in more places than just WP:NOR (for example, the appropriate and inappropriate uses of primary source material are currently discussed at WP:NPOV, WP:RS, and at WP:N... and are hinted at at WP:FRINGE and other guidelines).

What I would like to suggest is that we need multiple short "PSTS" sections ... a NOR/PSTS section that would focus on OR issues... a NPOV/PSTS section that would focus on POV issues, an RS/PSTS section that focuses on reliability issues. All of these would link to a new broad concept summary guideline... that would explain what the different source types are (including how different academic fields define the terms)... and summarize the various policy and guideline statements.

For this policy, it would mean the PSTS section would focus less on defining the terms (that would be done elsewhere) and more on how OR can result from misuse of the various types ... for some of the other policies and guidelines it might mean beafing up existing sections, or even creating a new one ... and of course we would have to write the summary guideline.

In other words... don't remove WP:PSTS... spread it out to several locations as appropriate. thoughts? Blueboar (talk) 05:03, 15 January 2009 (UTC)

OK, how can OR result from the misuse of the various types of source? Seems to me OR results whenever a source is used to introduce a new idea that does not appear in that source. (If that new idea does appear elsewhere, in some reliable but uncited source, the real problem is one of carelessness since citing that other source would make the material fully adherent to the policies.) Isn't this a simple situation being made complex by the very effort of trying to meticulously define primary sources (which definition depends in part on the use made of the source) and then specifying what can and cannot be done (with the cannot being used to forbid usages that result in OR.)

Were we working on the article on mutations I doubt we'd try to meticulously list every mutation - and we'd be correct to avoid that. Instead the article would discuss what is common to all mutations and perhaps use some few as examples, never attempting to make a complete list. Why torture ourselves by trying to do something even harder in policy articles? Minasbeede (talk) 12:05, 15 January 2009 (UTC)

Exactly! The whole section could be summed up as follows and nothing of direct relevance to the policy would be lost:

"Articles may include analytic, synthetic, interpretive, explanatory, or evaluative claims about published material so long as they have been published by a reliable source. For example, an article about a novel may cite passages from the novel to describe the plot, but any interpretation of those passages needs another source."

The PSTS section should perhaps be moved to WP:NPOV, since that is where it seems it would be the most relevant IMO. --Phenylalanine (talk) 12:51, 15 January 2009 (UTC)

To be honest I'd scrap PSTS as an ill-advised attempt to codify something that should be rather simple. Results of this ill-advised codification include debates about whether academic journal articles are primary or secondary sources, suggestions that review articles are "better" than new research articles on a subject (review articles are convenient, but represent the author's opinions and I've never seen critical responses to a review artcile, so IMO they are less well scrutinised in the medium term).

It would be much simpler to say e.g.

When using any source all you can do is summarise or quote it. You are not allowed to add any comment, interpretation, assessment or criticism of your own. If you think interpretation, assessment or criticism are needed, you must find additional sources that comment on the original source or the ideas or other content it presents, and where it is clear beyond reasonable doubt that they comment on the original source or its content. --Philcha (talk) 13:12, 15 January 2009 (UTC)

I don't think so. I might say "the Battle of Waterloo was a hard fought and costly battle." I can say that by citing several secondary sources, even if they do not use those precise words and I have to combine a couple, to prove the point (don't bother with SYN because it if definitely not a novel novel conclusion). To say the same thing using primary sources that do not use those precise words and needing to combine them to say such a thing is a breach of WP:PSTS. --PBS (talk) 14:48, 15 January 2009 (UTC)

IIRC Wellington said Waterloo was "a damned close run thing". If another commmander involved said something very similar, these woudl be statements from acknowledged experts and I wouldn't worry about PSTS any more than I would worry about 2 recent scientific journal articles advancing the same theory on a science subject - and we already know how PSTS gets itself in knots over science journal articles. Of course if another acknowledged expert published a different conclusion we'd be obliged to report that too, under WP:UNDUE. PSTS is too complicated for its own good. What really matters is the difference between simple, literal-minded summarisation (allowed) and interpretation / commentary / criticism by WP editors without support from reputable sources (not allowed). If there's a dispute, PSTS gets editors into theological wrangles about when a source is secondary or primary (which can often vary depending on the context). I think what I suggested would enable editors to resolve disputes by asking "What words in what source(s) support the words in the article?". --Philcha (talk) 15:14, 15 January 2009 (UTC)

I have to agree with Philcha here... if PSTS is somehow barring us from citing the Duke of Wellington for the statement "the Battle of Waterloo was a hard fought and costly battle" then there is a serious flaw with PSTS, not the citation. PSTS has taken on a life of its own... beyond its original intent of saying that we should not make WIKIPEDIA a primary source for information. Blueboar (talk) 15:25, 15 January 2009 (UTC)

Perhaps I'm wrong but my impression is that the reason for PSTS arises in situations like this. Suppose several secondary sources have concluded that the Battle of Waterloo was more of a skirmish and overall was minor. Then someone finds the Wellington quote and puts that into the Wikipedia article, indicating that Wellington, a key participant, thought the battle was hard fought. The PSTS advocates would (it seems to me) call that "improper use of a primary source" because the primary source indicates something other than what the secondary sources have concluded. One of them would summarily remove the Wellington quote. In the past it was indicated that PSTS arose from the field of historiography (that is, editors active in writing such articles inserted PSTS into WP:NOR.) The idea seems to be that even if Wellington thinks the battle was hard fought subsequent writings by authors of secondary sources take precedence over anything he wrote or said - and the editors who want PSTS as part of WP:NOR feel fully justified in removing any reference to Wellington's contrary opinion (again, this is an assumed example) and will instantly do so. They want PSTS to be a part of policy so that they can cite PSTS as justification for their removal. Minasbeede (talk) 15:30, 15 January 2009 (UTC)

Given your senario, I would say that removal on PSTS grounds would be a misuse of the policy. Blueboar (talk) 15:42, 15 January 2009 (UTC)

Re Wikipedia itself and similar sources, I think we should probably have a clause somewhere that specifically forbids the use of wikis (and their clones), with the partial exception of those where editing is restricted to acknowledged experts in their fields (I saw one about a month ago, of course I can't remember its name). Because of that experts-only wiki I wouldn't put too much emphasis on WP:SPS, although with any SPS I'd look for other sources in order to avoid problems with WP:UNDUE.

Re the hypothetical disagreement about the significance of Waterloo between Wellington and historians, the historians are greater experts on history, although I might be interested in why there was a difference of views. Re whether it was hard-fought, I'd more probably go with Wellington's view unless there was good reason to suspect him of hyping. That's partly based on this page by a renowned chess historian, which says historians should do history and players should do analysis of play. --Philcha (talk) 16:53, 15 January 2009 (UTC)

Just a note that you're probably referring to Scholarpedia. Looie496 (talk) 21:07, 15 January 2009 (UTC)

I deliberately choose my words carefully. Wellington did not say it was hard fought he said it was "the nearest-run thing you ever saw in your life." The point I was making is one can amalgamate and summarise several secondary sources into one or more sentences providing they do not contradict each other as that is part of the editorial process. What one can not do is amalgamate and summarise several primary sources even if they do not contradict each other, because it is original research, for which historians get paid and we can not do. For example: "The RAF raid to show the Russians when they arrive in Dresden what Bomber Command could do, started at 22:14 on 13 February " (1st primary source: an RAF readout at the briefing shortly before the attack, 2nd primary source: master bomber's report) The sources do not contradict each other, but taking that from primary sources is original research and in this case distorting the major reasons for the raid. --PBS (talk) 12:47, 16 January 2009 (UTC)

<outdent> I think the general definition of what primary, secondary and tertiary sources means, with examples, should be consistently worded in all the guidelines and policies where it's relevant. Keeping them consistent can be tricked out with a template or something so that the same words appear in all policy/guideline pages. I think it's more helpful that it appears in the policy/guideline page than it would be simply posting a wikilink to it. I don't see a value in page hopping to find out what a primary source means. I disagree with removing the PSTS in the policy altogether. I do see value in emphasizing that intepretations come from secondary sources, not wikipedians, and that wikipedia is not the place for users to analyze, synthesize, scrutinize, bastardize, polemicize or any otherwize interpretize primary source materials. That's what secondary sources are for. :0) Professor marginalia (talk) 18:04, 15 January 2009 (UTC)

Silly Question

Why is wikipedia concerned with verifying information by citing mainstream sources of information that may be and in many cases are susceptible to bias, and not concerned with truth? This is an encyclopedia, not a mirror of mainstream information sources. Has this fundamental issue been raised before, if so where can I find the discussion? And why was the emphasis not changed? If not, I'd like to discuss why wikipedia is not concerned with truth, yet claims to be an encyclopedia. It should also be noted that WP actually allows original research in images. Nick carson (talk) 03:26, 14 January 2009 (UTC)

WP:V covers this. --Ronz (talk) 03:55, 14 January 2009 (UTC)

It isn't that we don't care about "truth"... it's that "truth" is subjective... there is often a legitimate debate as to what the "truth" actually is. Thus, we use use a different standard... verifiablility. If something is both verifiable and true... so much the better. Blueboar (talk) 06:09, 14 January 2009 (UTC)

In general, sources that can reasonably be regarded as reliable don't use the word "truth" to describe what they're asserting. Why should an encyclopedia such as WP assert such a thing as "truth"? Does any remotely credible encyclopedia assert it's putting forth "The Truth"? Even the Catholic Encyclopedia asserts no such thing. Rather, encyclopedias tend to say they're putting forth information of some kind. Did I miss anything important? ... Kenosis (talk) 06:57, 14 January 2009 (UTC)

I speak of truth in terms of reality, fact, etc, such as the 'sky is blue' statement. Many of us working with detailed or obscure information that may be significant and noteworthy as per WP policy, are forced to omit such information fundamentally because it is detailed or obscure or unknown by mainstream, verifiable sources of information as per WP:V.

There are alot of things that are true, yet are not verifiable as per WP:V, surely we don't wish to omit crucial, true information from WP. Need we mention the tendency of WP to favour mainstream points of view, a contradiction as it conflicts with WP:NPOV, and thus the omission of significant, notable, important information that may also be obscure or unknown by many.

I simply can't agree that sources of information subject to bias and lack of quality constitute a better standard than original research encompassing multiple sources of information which may or may not include direct factual observation. Nick carson (talk) 11:33, 14 January 2009 (UTC)

I think I understand the puzzlement behind the original question. The "verifiability" bit is largely to prevent people from pushing their own personal points of view - a problem which often arises in an open encyclopedia, but which can usually be dealt with by citing WP:NPOV, WP:V, WP:RS etc. and deleting the offendinging content.

But WP:RS sets what it regards as minimum standards for reliability, and admits that sources that meet these criteria may not be reliable in particular cases. For example a scientific article several years ago by a reputable author in a reputable journal may have been refuted in the meantime, so you could either ignore that theory or cite the later refutation and any theory that is now regarded as "true" or even promising, provided you don't over-hype the later theory. In practice it's mainly just common sense, and the rules only get an airing when there's a dispute. --Philcha (talk) 11:40, 14 January 2009 (UTC)

I think Nick carson's later post (11:33, 14 January 2009) goes a bit too far. "The sky is blue" is generally an uncontroversial statment and does not need refs - except in some special contexts, e.g why it's blue or how it looks to people with certain colour-vision deficiencies. Re "sources of information subject to bias and lack of quality", the formal, usually process-/organisation-based criteria at WP:RS are only a start, and relevant cricisms of one WP:RS by another WP:RS are fair game. However there is no conflict between WP:NPOV and WP:V, as WP:NPOV stresses the need for WP:RS. These policies are needed to prevent WP from being deluged by WP:FRINGE theories and their enthusiasts. --Philcha (talk) 11:51, 14 January 2009 (UTC)

Grouping non-mainstream subject matter into WP:FRINGE isn't fair. Just because a theory, musical act, town, suburb, river, etc isn't well published by a RS and notable within the mainstream doesn't mean that it isn't important or notable. Granted, I imagine there are alot of people out there who push untrue, fundamentally fictional, fringe theories and subject matter, but what about genuine subject matter that may only be known directly by a 'fringe' minority that may not publish their information in the conventional sense? Such an example would be a piece of artwork that influences a group of artists that subsequently influence mainstream popular culture at some point down the road. Another example would be a musical act who's work in melding particular styles/genres is genuinely original and is later picked up by another mainstream popular musical act who then gain credit as they are published in the conventional sense, years down the track? WP policy has progressed to encompass reliable blogs, what about reliable zines, street press, etc? Why can't we progress now?

Not only this, but I know of several editors and estimate many thousands, who spend a significant portion of their time contributing to WP trying to cite resources for information that they could otherwise prove and verify with OR or direct observation. Surely in such a case as OR and direct observation are used, the information can still be challenged as in other WP policies, and if it is a genuinely false fringe theory or subject matter, the false information will inevitably be discovered and corrected. Sorry for the passionate diction but I'm baffled at how things such as OR have become somewhat of a dirty word on WP, there are just as many arguments for why they can be used as a progression of current WP policy and verified in the same manner. Nick carson (talk) 12:22, 14 January 2009 (UTC

It is very simple. Wikipedia, by design, is not the place for original ideas or results to appear. There are many venues where original ideas and results can appear and if these are suitably notable then the ideas and results reported elsewhere can appear in Wikipedia. Therefore, anyone with a great new idea should automatically not think of Wikipedia when considering where to first publish it. The restriction has nothing to do with the utility or merit of the material that is forbidden, it's simply a chosen policy. Allowing only already-published material to be used as a basis for Wikipedia greatly strengthens Wikipedia, to the extent that the restriction is honored.

You didn't say it but you could have that the major print encyclopedia, the Encyclopedia Britannica, has often times solicited major articles from world experts (Albert Einstein, for example.) In all probability such experts have, in the material they contributed, included some original research - and that original research enhanced the value of the contribution. Wikipedia works in a different way and has, in effect, chosen to forbid such enhancement. Given the open nature of Wikipedia that is a good choice.

Nothing I say is meant to deprecate the value of the sort of research you appear to desire to be done. The exclusion of OR isn't based on value or lack thereof. If you want to do original research, do it. Just don't publish it first in Wikipedia. Minasbeede (talk) 01:16, 15 January 2009 (UTC)

In the interests of encyclopedic geekiness, are you aware that "the sky is blue" is challenged (in a way) by the article: Distinguishing blue from green in language

However, on a more serious note, if something is not verifiable, then it is unfair on a reader to state it as a fact, so we have agreed that it is outside our mission. For example, a newspaper might say that "Chocolate mousse is a delicious, satisfying snack", and a reader will assume that is the opinion of the food critic, or may think it to be a fact. But if we say anything at all, we should only say "Chocolate mousse is marketed as a snack food, that the Boggington Chronicle says is both satisfying and tastes good." At least that is my simple-minded way of viewing verifiability.

Nick, could you post examples of facts that you would like to put in Wikipedia but cannot as they are unverifiable?

--Hroðulf (or Hrothulf) (Talk) 12:24, 14 January 2009 (UTC)

By the same token, it is unfair to the reader if a verifiable piece of information was cited from a source that favours mainstream POV, let alone anything that is not fact as synthesised through OR or other means. If such information is outside of our 'mission' then why not increase the 'mission'? I merely ask, why not?

I can, one can be found in the band "Ohana", who have incorporated in their work elements of minimalism, math rock, post-hardcore and post-punk to create an original sound that is as yet unheard of in Australia at least. Such information can be ascertained by direct observation by people with knowledge and experience in such areas and has thus been published in countless zines and street press articles. One of the more 'mainstream' sources being several issues of the magazine "Mess+Noise".

Another example can be found in tracing fashion trends from sub- and alternative cultures that permeate into mainstream/popular society. Such information that is currently not included in WP but is well known to many and well published by reliable unconventional sources.

However, this example is not my reasoning for attempting to generate discussion on the topic. I'm looking for some in depth discussion into how we can improve the core WP policies by encouraging a progression of their contents and goals and introducing new ideas that can increase the inclusion of notable and specific information on WP, the quality of articles in general, the inclusion of OR assessed on the individual's knowledge, experience in the field and the reliability of the sources they are working with, and to decrease WPs reliance on mainstream sources of information by making WP policy inclusive of unconventional sources such as zines, blogs, street press and other alternative forms of information conveyance, not just published, but synthesised and collated. For what more is WP without such things than a mere reflection of mainstream, popular, conventional society? There is alot more to life than that and to exclude such information from an encyclopedia is to ignore a significant portions of one's own history and present. Nick carson (talk) 12:40, 14 January 2009 (UTC)

Hroðulf, nice catch re Distinguishing blue from green in language!

Nick carson's comments (12:22, 14 January 2009) highlight that "mainstream" is ambiguous. I was thinking mainly of science because that's mainly where I edit, but in the arts, including popular culture, a lot of significant ideas start as "non-mainstream". However the next sentence of his comment shows the problem with "non-mainstream" anything: if no-one outside those directly involved pays attention to a theory, musical act, town, suburb, river, etc., then by real-world standards as well as WP:NOTABILITY it isn't notable. Without this rule WP would simply be the world's biggest free advertising billboard - and widely ignored, as very few people deliberately look at ads (yes, I can provide refs for this).

Re Nick carson's examples of innovative art, music, etc., it's not WP's job to predict future trends - an activity in which the success rate is rather low. Per the 2nd Sturgeon's Law I would guess that the great majority of innovative art, music, etc. is soon and in most cases deservedly forgotten. Almost by definition the ones that succeed and are notable in the real-world sense are those that are noticed by "mainstream" commentators.

Re the difficulty of sourcing material that's widely accepted but not obvious to primary school kids, I sympathise. In science articles, editors often need to obtain decent text-books to source all the basic stuff that's not discussed in academic journals - public libraries and Google Books have been good to me. Since my interests are not confined to science, I realise that it's harder to find text-books on non-academic subjects - but Google Books is still worth a try. If you're that keen on the subject, you'll make the effort to find sources. There are other methods - asking at Wikiproject pages, asking specific editors who are active in that area, finding enthusiast societies and asking them (that's how I got info on the contributions of Adolf Anderssen to chess problem composition). --Philcha (talk) 13:11, 14 January 2009 (UTC)

Folks, I think we have ventured well away from the purpose of this talk page... which is to discuss the NOR policy. I probably started us down the road in the first place, by attempting to be nice and answering Nick's question... and I now appoligize for that. Ronz had it right to begin with... this is a discussion that should be taking place at WT:V, WT:FRINGE and perhaps several other guideline and policy talk pages. It isn't, however, a discussion that we should have on this talk page... it does not involve WP:NOR. Blueboar (talk) 14:45, 14 January 2009 (UTC)

Perhaps we should have moved it to WT:V at the outset, but since we didn't, it is ok to have an off-topic discussion if it is an answer to a question that the original poster thought was on topic. The original post was not spam, which should be the only reason to kill a thread based on topic.

Here's my attempt to draw this to a close: Nick Carson wrote:"why not increase the 'mission'". Wikis can of course be more flexible about verifiability and truth, to investigate non-mainstream parts of human knowledge - that is one reason why there are many other wikis, such as those at Wikia.

--Hroðulf (or Hrothulf) (Talk) 14:57, 14 January 2009 (UTC)

<- I'm sorry, but the mere use of the word "truth" as a proposed standard for writing an encyclopedia is a total red herring. As opposed to what? Editors are lying? Engaging in a conspiracy to selectively mislead WP's readership? I think not. Or is the assertion that the lack of an official standard of "truth" results in an unnecessarily high rate of untruthful information? No, it's a red herring, a phantom. WP already deals with this issue by its three existing content policies, WP:NPOV, WP:V and WP:NOR. ... Kenosis (talk) 19:37, 14 January 2009 (UTC)

The statement that no one pays attention to to a theory, musical act, town, etc who is not directly involved is unsubstantiated. And even if it was, attention is no grounds for classification of notability or importance. Please explain what you mean by 'real-world standards', I hope you don't mean mainstream/popular society standards. Success and notability regarded by mainstream commentators is no indication of a subject matter's, particularly in the case of music, art, film, etc, notability or importance. Note that my use of the word 'truth' stems from the fact that it is used in this policy itself.

My words are being taken somewhat out of context, some examples... I never said it was WPs job to predict future trends. I never suggested that WP editors were lying, nor did I suggest they were engaged in a conspiracy, let alone did I even introduce such words into the discussion... Please understand what I wrote, re-read it if you like and absorb it. Other unsubstantiated and uneducated comments were made including... Innovative art, music, etc is 'deservedly forgotten'. The assertion that these discussions belong in WT:FRINGE.

Philcha, I guess part of my argument centres around the fact that one or a number of sources used to cite particulars and prove their verifiability is not as reliable as a synthesis and collation of a collection of sources, as conducted by one or more or an entire team of editors, as is the case with many article re-writes and wikiprojects.

In regards to Blueboar's suggestions, I totally understand but have in the past been passed around the 3 core policy talk pages, where is the discussion be had regarding the progression of these policies if not on their talk pages? Am I to copy and paste discussion so that it appears on each talk page simultaneously?

In regards to Hrothulf, that may very well be what inevitably happens to WP, editors may move to more specific Wikis which will ultimately be disjointed and separate. Why deal with detailed and specific information by palming it off to all sorts of sub-wikis? Why not keep it within it's original context, located within a centralised source such as WP itself. Nick carson (talk) 23:50, 14 January 2009 (UTC)

Nick, your question was, "Has this fundamental issue been raised before, if so where can I find the discussion?" The policy against original research is one of the few "fundamental" rules in wikipedia-goes way, way back. (2004?) I don't have a crystal ball, but I think it's safe to say the odds of this policy being reversed are nill. If you go to the bottom of the main policy page you'll find there some links to some of the earlier discussions about it by Jimmy Wales. Professor marginalia (talk) 00:06, 15 January 2009 (UTC)

I never requested or suggested the policy be reversed, I'm initiating discussion as to how we can improve upon it. Early comments by Jimmy Wales helped to establish the current versions of the core policies which as you said, perhaps go back to 2004, I'm looking for discussion that is a bit more recent than 4 or 5 years ago. Why are contributors to policy talk pages more keen to set the policies in stone than they are to keep them progressing, evolving, encompassing? Perhaps this is a case for WP:IAR in that the core policies are not willing to be amended and improved upon thus preventing editors from improving and maintaining WP. Nick carson (talk) 00:14, 15 January 2009 (UTC)

With Wikipedia is not a publisher of original thought and Wikipedia is not an anarchy in effect, Ignore All Rules would be a futile waste of time. Start with an essay and see it you can inspire some kind of a grassroots movement among editors interested in changing the policy. Professor marginalia (talk) 00:28, 15 January 2009 (UTC)

<-- The reason for the fundamental policy mandate which says "Verifiability, not truth" is, I suspect, that the word "truth" lies somewhere in the area of both useless and meaningless as an editorial policy. From the perspective of WP readers who may wish to question anything in WP, the only practical question of any relevance when someone writes something in WP that the person writing it asserts to be "truth" is, essentially, "where can I, the reader, double-check this?" Without a doubt there are things-in-the-world that are truthful but not independently verifiable, but the mandate for us, which began to be set in place as early as 2003, was that WP should be limited to things that are verifiable in the event any question arises about the veracity of a statement. See, e.g. the earliest version of WP:V. See also, e.g., this email from WP founder Jimmy Wales, which explains why the relevant standard is "verifiability, not truth". ... Kenosis (talk) 04:44, 15 January 2009 (UTC)

Such a mandate is outdated as it excludes information which is "true" yet not independently verifiable, thus excluding crucial information from inclusion in WP. My argument is that there is a better way than mere verifiability and better goals than merely the inclusion of only verifiable information. In the meantime, or in the event that editors are reluctant to amend and progress such policies, such exclusions render WP unreliable as a result of its uncomprehensive nature. Professor marginalia, I may just take up your advice, though I wish there was 10 lives I could live to create the time needed to devote to the many things that I feel are worth devoting my time to. I fear much of this discussion has glazed over a portion of its contributors whom have subsequently resorted to policy reiteration. Nick carson (talk) 13:03, 15 January 2009 (UTC)

Nick, turn your complaint around and you can see the flaw... How do we know whether information which isn't independently verifiable actually is "true"? You say that "crucial" information is being excluded... can you give us an example? Blueboar (talk) 20:03, 15 January 2009 (UTC)

Nick carson, the only actual example you've mentioned is a band called "Ohana". I Googled for "Ohana" and for "Ohana band" and the only hit I got that looked relevant was an entry on MySpace. I guess many WP editors get a little frustrated about the rest fo the world's lack of interest in some of their favourite topics. Hell, I could re-write Fermi Paradox right now to include my views on the serious omissions in all the analyses I've seen - and it wouldn't last 5 minutes, because I can't cite sources (apart from that ultimate authority, me). So instead I edit articles on subjects for which I can find sources. --Philcha (talk) 20:34, 15 January 2009 (UTC)

It's a bit sad. What is holding WP back from becoming more detailed, specific, encompassing, inclusive and less reliant on mainstream, conventional sources of information? I'd like to see some comments contributed regarding the rather than poke holes in my example and keep telling me that WP policy is holier than the bible and will never progress! Nick carson (talk) 12:14, 17 January 2009 (UTC)

Wikipedia is not meant to be a fully comprehensive compendium of everything everyone has thought about a particular subject. Without its policies on verifiability and reliable sources it would, at least for many, lost what I see as its heart.

Edit warring

Would the people who are edit warring over the inclusion or non-inclusion of various phrases in WP:PSTS please knock it off already! Add a {{dubious}} tag if it contains stuff you don't think it should, or remind yourself about WP:The Wrong Version, but please stop.

Thanks, WhatamIdoing (talk) 05:54, 15 January 2009 (UTC)

I would also point out that there seems to be growing consensus that PSTS may need to be majorly rethought ... so edit warring over the current language is a bit pointless. Blueboar (talk) 15:30, 15 January 2009 (UTC)

How about throwing the whole thing into a volcano, as a sacrifice. Problem is, these sorts of sacrifices tend to bring only temporary relief, and after awhile another sacrifice is needed to keep the gods happy. ... Kenosis (talk) 18:02, 15 January 2009 (UTC)

Let's rethink first about the purpose of PSTS: is it to provide needed guidance to editors or is it to create a policy clause that enables instant deletion of material characterized as being inappropriate use of primary source material? As noted above, my notion/impression is that PSTS is there to prevent editors from citing primary sources that disagree with secondary sources on the same issue and to give license to editors to cite PSTS as grounds for removal of such material. I have moderated in my view: maybe the removal is righteous (in Wikipedia policy terms.) I can't say/don't know. The OR which seems to be forbidden consists of consulting a primary source and finding it differs from secondary sources. I can easily believe the secondary sources are generally better thought out (or substitute a better term than "thought out" here) than are primary sources but that word "generally" may indicate a problem (since policies aren't for the general case, they are universal.) Minasbeede (talk) 18:33, 15 January 2009 (UTC)

What is needed is some sort of "constitutional convention" concerning source-typing issues in WP:NOR, WP:N, WP:P, and WP:RS, which would be a sort of re-boot, perhaps moderated by a respected Wikipedian such as someone from the Arbitration Committee. Everyone who participates in the process would have to agree that they would abide by any consensus achieved during the process and not try to either "change it back" or to reintroduce something for which consensus was not established during the process. Nothing gets back into the draft unless there has been an established consensus. COGDEN 19:22, 15 January 2009 (UTC)

Hey, let's demand proof of consensus, and lacking a clear demonstration of such to the satisfaction of all critics of the policy, just delete it, maybe make a guideline or an essay out of it. ... Kenosis (talk) 19:42, 15 January 2009 (UTC)

Rescind all the existing policies, and engage in a community wide "constitutional convention" presided over by the Arbcom? -- Goodgahdahmitey, I think I need a drink. Professor marginalia (talk) 20:20, 15 January 2009 (UTC)

I disagree that we need to "think first about the purpose of PSTS". That's what has got us into this mess. Get an independent and sourced PSTS definition up, with examples. Then we can discuss how policy and guidelines fit around them. For example, I see two separate sourcing issues:

A primary source used in WP text that is discussing that source. For example, the novel as a source for the article on the novel. There's a great deal of consensus about this and when we need secondary sources. We are currently quite firm about this.
A primary source used in WP text about a topic other than the source. For example, citing a diary vs citing a biography in an article on the person. Or citing a randomised controlled trial vs citing a review in an article on a drug therapy. In this case, there's less consensus and our guidelines reflect that one form of source tends to be better than another, etc. We probably can't come up with firm rules but may develop guidelines, some of which might be topic-specific.

A proper discussion on our application of PSTS can only occur once we all have a common definition. As long as some of us say "I don't want that to be a primary source, because it affects my argument" then we will get nowhere. Colin°^Talk 21:40, 15 January 2009 (UTC)

Colin... If a policy statement has no purpose (as some are arguing), it shouldn't be in the policy in the first place. And if it does have a purpose (as others argue), surely we need to at least reach agreement on what that purpose is before we address the issue of what it says. Blueboar (talk) 22:32, 15 January 2009 (UTC)

I'm not talking about the "policy statements". I'm talking about definitions. There are too many places (as noted) where policy and guidelines mention primary and secondary for us to leave them undefined or inadequately defined (as is at present). Now, if we decide that certain restrictions are independent of primary vs secondary issues, then we can state them without referring to primary and secondary. But I think we'll find that some (both hard restrictions and soft guidance) will benefit from using those terms. So lets agree on the definition of those terms and then discuss the "policy statements". And I agree that purpose comes before text. All I want is that definitions are independent of Wikipedia and its peculiar restrictions on the use of sources. Colin°^Talk 23:28, 15 January 2009 (UTC)

Minasbeede's "Let's rethink first about the purpose of PSTS: is it to provide needed guidance to editors or is it to create a policy clause that enables instant deletion of material characterized as being inappropriate use of primary source material?" (18:33, 15 January 2009) gets close to the heart of the matter. I'll be blunter: are we trying to guide editors, or to provide grounds for wiki-lawyering? I think the current PSTS is good only for wiki-lawyering. In a previous thread I suggested an alternative guide that focussed on the principle of WP:NOR - in sumary, "interpretations / assessments / criticisms based only on WP editors' own opinions are not allowed" - which I think is fairly easy to understand, even for newbie editors. --Philcha (talk) 22:40, 15 January 2009 (UTC)

The purpose of PSTS should be to define PSTS. What we say about NOR may or may not refer to primary vs secondary, but that should be free for discussion. We need to stop linking policy rules with definitions. Colin°^Talk 23:28, 15 January 2009 (UTC)

User:Minasbeede presents a false dichotomy. There is a middle ground between providing guidance and enabling instant deletion. We don't need to argue which end of the spectrum is better than the other. We need to find a sensible solution that meets the usual overall goals of all Wikipedia's policies, which are to simultaneously provide:

* general principles AND

* education for those unfamiliar with the concepts AND

* enough detail that people of good will don't end up with radically different perspectives AND

* room for editorial judgment and common sense.

There's nothing in any policy except BLP that mandates the instant deletion of disputed material. Let's not pretend that PSTS will do that, either. WhatamIdoing (talk) 00:24, 16 January 2009 (UTC)

While I could argue some of your points that seems unwise, given how long this has been discussed/argued (enough, already.) I say "Great! find that middle ground!" I like your four bulleted points. I'd say you've made a useful positive contribution.

Any more from me would probably be argument masquerading as something else. Not needed. Minasbeede (talk) 00:31, 16 January 2009 (UTC)

I'm not sure about "Let's not pretend that PSTS will do that (mandate the instant deletion of disputed material)". Like any yother rule related to WP:V, WP:NOR and WP:RS, PSTS will enable editors to claim that others' edits are inadequately sourced, and deletion is permitted in these circumstances. The sad thing is that some editors use the rules to harass other editors, which is why wiki-lawyers are are unloved as other types odf lawyer. I support the principles of WP:V, [WP:NOR]] and, with more reservations about the details, WP:RS. The amount of debate recently about PSTS, which I did not start but in which I've participated, indicates that PSTS is complicated and difficult to interpret, i.e. wiki-lawyers' heaven. I suggest it would be better to replace PSTS with a statement of what it treisd to achieve and what it tries ot prevent, in language that newbies can understand.

If we could do that, I'd also support any move to produce a simple, 1-page summary of WP:V, WP:NOR and WP:RS that we can issues to new editors, and perhaps have incorporated into the mastheads of Talk pages (except User Talk, where WP:CIV and WP:DE are all that really matters).

This is not bleeding-heart liberalism on my part. The NY Times and The Economist have both written about the harm done by wiki-lawyers. The NY Times piece is a commentary on a book that someone's written about how to avoid trouble with wiki-lawyers. --Philcha (talk) 06:29, 16 January 2009 (UTC)

(outdent) I think everyone (except wiki-lawyers) recognizes that the PSTS section devolved out of an accident, and that it now dilutes and corrupts NOR policy's real message. As the numerous discussions make clear, the section is bad, but a {{contested}} tag voids the whole policy, and we can't have that.
Instead, we (effectively) ought to really throw PSTS "into a volcano", by which I mean move it. Here is why:

The PSTS section is based on false premises, and, in its present form derives from a version written by people whose first encounter with the term "original research" is this page itself. By adding two plus two, those people vaguely "knew" that its bad for the 'pedia, but do not really comprehend why, nor did they really understand why NOR policy had the phrase "Wikipedia is not a primary source" to begin with. The present PSTS is a product of those unfavorable circumstances, and the lack of comprehension is (for example) still evident in #1) the specious premise that primary sources are especially subject to OR, and #2) the sui generis premise that there are different definitions of p/s/t sources, and #3) that the 'pedia has a need to adopt one that is different from that used elsewhere.
These erroneous premises aside, in its conclusion the PSTS states that "deciding whether primary, secondary or tertiary sources are more suitable" is a matter "of common sense and editorial judgment." This is an explicit declaration of what all the 'should's of PSTS already imply, i.e. that PSTS is a guideline and not policy. And this license to interpret PSTS any way one wants is deeply ironic. After all, interpretation is what NOR policy is supposed to prevent.
Whatever the PSTS section's redeeming features (if any), a p/s/t distinction has nothing to do with NOR policy, which is simply stick to the source. And that means any source. Whether that source is primary, secondary or bazzilionary does not affect whether that source is accurately parroted. "Original research" is the production of something new, and it does not matter what that new content is based on. It can be anything.
The rationale for NOR policy is this:
Because Wikipedia is supposed to be a reference work, its articles cannot contain ideas that have not been expressed before. In contrast to academic research (where new ideas based on older ones are welcome and are known as "Original Research") inferences may not be made in a reference work.
The explanation of this basic tenet originally used the terms "primary"/"secondary" etc. But it is not necessary to use p/s/t to express that idea, and further, the present-day PSTS no longer reflects it.

So... since PSTS tells us that the section is a guideline, then it belongs in/on a guideline page and not willy-nilly stuffed into a policy where it contributes nothing and only serves as a distraction. Once moved, it will cease to compromise NOR and RS policies. If PSTS's high priests continue to think that it is policy material, then they can RFC it as such. Either way, PSTS's rape of WP:NOR (and, as we see, of WT:NOR) has to end. -- Fullstop (talk) 16:53, 16 January 2009 (UTC)

I would be much happier with the world if PSTS were a guideline instead of part of a major content policy. Phil Sandifer (talk) 16:54, 16 January 2009 (UTC)

Fullstop, congratulations on your elegant and comprehensive demolition of PSTS! --Philcha (talk) 17:08, 16 January 2009 (UTC)

WP:PSTS's rape ? of WP:NOR? LOL. I really think it might be time to throw the whole section into the volcano. I'll start a proposal section below. ... Kenosis (talk) 17:47, 16 January 2009 (UTC)

Suggestion: The debate between the "PSTS inclusionists" and "PSTS deletionists" has gone round in circles for two years at least. It's not going to be settled in a "don't edit war" thread. And it won't ever be settled unless there's a clear choice put to us, with ample notice given the community, and clear evidence of the "community will" given. Anybody remember the WP:ATT trainwreck? Lotta hard work, lotta wasted time, lotta complaining when it was finished, only to be reduced to "an essay". This is a pillar policy. Realistically, a big change to it will require a huge amount of work behind it to give it any chance of "sticking". The battle between "PSTS inclusionists" and "PSTS deletionists" won't end any differently than before riding same merry-go-round-and-round like we're doing now. Professor marginalia (talk) 17:19, 16 January 2009 (UTC)
Fantasy visions of "PSTS inclusionists" and "PSTS deletionists" aside, a suggestion to "do nothing" is not a suggestion. Besides, I didn't say it ought to be deleted. I said move it. Let the high priests of PSTS put their money where their mouth is. They have nothing to lose if the religion that they are preaching is the right one. Their claim that its god given (its been there forever) is not a rational argument in an enlightened world where the real meaning of "primary"/"secondary" is actually known, where the orthogonal relationship between these and NOR policy is actually understood, and where PSTS is recognized for the irrelevance that it is. -- Fullstop (talk) 18:05, 16 January 2009 (UTC)

Returning to regularly scheduled programming: Consensus is that the section sucks. The ball has been in the PSTS-high-priest's court for ages, and the offer to move is their last chance, playing by their own WP:POINTy rules. If they refuse to take it, then point-match-game, and the ring-a-roses is over. Goodbye idiocy, sanity prevails. Enough of the filibustering! -- Fullstop (talk) 18:05, 16 January 2009 (UTC)

The "last chance" lol? Maybe the "edit warring" thread is an apt place to discuss the proposal after all. Professor marginalia (talk) 18:27, 16 January 2009 (UTC)

Prof, was your edit summary "banners all bravely unfurled" statement of WikiCreature allegiance :-) --Philcha (talk) 18:35, 16 January 2009 (UTC)

That's one interpretation :) Professor marginalia (talk) 18:54, 16 January 2009 (UTC)

It was predictable that a demand to end the filibustering was responded to with another digression, which includes the usual failure to address the point, and going off on a tangent once again. Some people just don't get the fact that their opinion don't count when they don't express themselves. Attention silly rabbits! We will not follow you down your rabbit hole. If you can't stay around on topic, then we'll end up having tea without you. -- Fullstop (talk) 20:19, 16 January 2009 (UTC)

Here are my thoughts then. Don't edit war on the policy page. Waiting to hear yours with any relevance to the thread's topic. Professor marginalia (talk) 20:26, 16 January 2009 (UTC)

WP:POINT predicted that yet another feint would came along. The last one ironically about not following the thread. My comment was in fact a direct response to top-post's remark about the dubious tag, and why it is bad, and what should instead be done. The call for move may also be seen vis-a-vis WhatamIdoing's call for middle ground. My comment may also be seen in contrast to, say, gratuitous remarks such as "don't edit war on the policy page" while simultaneously refusing to acknowledge the problem that precipitates those edit wars.

A sustainable argument for keep has not once, ever, been put forward. Never. Instead, every comment that says don't keep is either followed by an off-topic nonsense, or no comment at all. With no sustainable arguments for keep, and no sustainable arguments against not-keep, there is only one consensus evident. And that is not-keep. You lose Professor, and that is because of inane comments that count for diddly. If you (plural) want to count, then get off that high horse and play ball. -- Fullstop (talk) 22:29, 16 January 2009 (UTC)

So let's play ball then. Anyone who says this thread titled "Edit warring" posted on the NOR talk page (which resulted from a series of reverts of a single phrase of the NOR policy) is the appropriate place to gather the solid and sustainable consensus needed for either a) trashing *four* policies, WP:NOR, WP:N, WP:P, and WP:RS outright and rebuilding from scratch or b) removing/relocating the PSTS clause alone, is either brand new here or grandstanding. Professor marginalia (talk) 00:14, 17 January 2009 (UTC)

Who's trying to build a case for trashing or re-building from scratch WP:NOR, WP:N, WP:P, and WP:RS? I've argued that PSTS is too complicated, a magnet for wiki-lawyers, confusing to most editors (frankly I read half, foundit unhelpful and went back to what I was doing)) and its aims could be achieved by a simple "editor's are not allowed to interpret sources". IMO that's an attempt to make WP:NOR effective without the confusion and ill-will that wiki-lawyering creates. --Philcha (talk) 12:35, 17 January 2009 (UTC)

"reboot" - I was being called in the carpet for not taking such battle cries posted in this thread seriously. Professor marginalia (talk) 17:39, 22 January 2009 (UTC)

I've left an invitation at the village pump for the broader community to weigh in below, so as to try to get either a clear sense for how much opposition there is to WP:PSTS as policy and/or to get some kind of basic affirmation of consensus in favor of it as part of WP:NOR.
..... It's at Wikipedia:Village_pump_(policy)#Proposal_to_remove_WP:PSTS_from_WP:NOR. Perhaps someone might like to also start a policy RfC to invite yet broader feedback so as to try to get a better handle on this. If it's true, as some have said, that consensus was never demonstrated for WP:PSTS as part of WP:NOR, then we ought be able to get a better sense of where it stands in the eyes of the broader community, rather than just among those who frequent this talk page. ... Kenosis (talk) 20:30, 16 January 2009 (UTC)

I agree about getting "a better sense of where it stands in the eyes of the broader community, rather than just among those who frequent this talk page". As soon as possible, or leave it for a day or two so that arguments on both / all sides of the issue are fully developed? --Philcha (talk) 21:38, 16 January 2009 (UTC)

Proposed removal of WP:PSTS from WP:NOR

Straw poll: In light of complaints about WP:PSTS and assertions by several users that PSTS should be removed from this policy, it might be useful to get an idea how many users are currently in favor of deleting WP:PSTS outright and placing the existing content on a new page that tentatively downgrades it to a guideline. ... Kenosis (talk) 17:47, 16 January 2009 (UTC)

Keep as part of WP:NOR. Despite vociferous complaints from some, the PSTS section plays a useful role as an editorial policy. ... Kenosis (talk) 17:47, 16 January 2009 (UTC)
Keep in NOR. But polling "those currently in attendance" on the talk page is a bare beginning. Professor marginalia (talk) 17:55, 16 January 2009 (UTC)
Delete - "editor's interpetations of sources are not allowed" is sufficient --Philcha (talk) 18:07, 16 January 2009 (UTC)
Delete per Pilcha, or Move and file RFC for elevation to policy status. Hypothetical and never rationalized notions of its "role as an editorial policy" contradict the section's own statement that it is A) a guideline and B) superseded by editorial decisions. As noted above, if there were an iota of policy material in the PSTS section, then the prophets of PSTS should put their money where their mouth is and make the contents of PSTS run the same course that turns wannabe-policies into policies. As such, the contents need to be on their own page, for which an RFC is then filed. If the RFC turns out as the prophets are convinced it will, then a transclusion into this one would be justified. Such a validation was never made, and needs to be done. -- Fullstop (talk) 18:19, 16 January 2009 (UTC)
Delete (actually, move to a separate page that is a guideline.) PSTS fails as a policy because it has proven impossible to find wording that does not provoke continued controversy. It also fails as policy because it consists of a strange mix: primary sources are allowed, but using primary sources can serve as justification for removal of an edit, often without any explanation at all on the relevant talk page. In other words, PSTS is so unclear as to be useless as policy and tends to be sometimes used in what appears to be an arbitrary manner. An alternative would be to alter the wording of PSTS so that it no longer provokes controversy. Blueboar has attempted to accomplish this for over a year and it hasn't happened so this alternative may be an impossibility. In effect, too, PSTS is an attempt to begin to enumerate the ways OR can be created. Such enumeration is an unnecessary and impossible task. As a guideline PSTS might assist editors in analyzing what they intend to include and determine whether it is OR, and as a guideline PSTS could be both detailed and open-ended, since as a guideline it does not have to be definitive. Policies should be definitive, or at least not ambiguous. Guidelines (it would seem) ought to cover all major points but if something is overlooked or left out that has no bad effect with respect to policy since a guideline is not a policy. When it started out back in 2005 PSTS encouraged the use of both primary and secondary sources. Editors should be cautious in how they use all sources; singling out primary sources as needing caution falsely paints such sources as being less desirable, seemingly in all cases. If the policy wording has the sense of "secondary sources are preferable most of the time" then the policy provides no useful guidance (and certainly no grounds for summary removal) since in all cases the policy is applied to specific articles - "most of the time" means that some of the time secondary sources are not preferred, which might easily be the case for any specific article. If the policy wording attempts to enumerate all the exceptions in which primary sources are preferred (or acceptable) then the policy wording will never be complete and the attempt to make it so will drag on for years. "Drag on for years" - sound familiar? Minasbeede (talk) 18:38, 16 January 2009 (UTC)
Keep It's an exceptionally useful part of WP:OR, helping guide editors on how to identify and resolve the most common OR/SYN problems. PSTS issues are indeed complex, and preventing OR/SYN problems is often difficult and controversial. Complexity is no reason to remove it, but an indication that it needs to be improved, as multiple editors have indicated their willingness to do. Wikipedia without controversy is impossible. Often controversy is a strong indication that we are improving Wikipedia. --Ronz (talk) 18:58, 16 January 2009 (UTC)
- You may (or may not) have a point. It may (or may not) have already been addressed in the last longer comment in the section above this one. I.e. the one before Kenosis and Professor marginalia attempted to hijack the discussion. The notion that PSTS is useful is voided by the fact that it explicitly says "do what you will". The notion that it "guide editors on how to identify and resolve the most common OR/SYN problems" seems to indicate guideline, and since p/s/t distinction has nothing whatsoever to do with OR or SYNTH, everyone here would surely appreciate an explanation of how such a thing might occur. -- Fullstop (talk) 20:29, 16 January 2009 (UTC)
Keep are you nuts? Semitransgenic (talk) 19:38, 16 January 2009 (UTC)
- Are you literate? If so, could you care to explain why anyone would/wouldn't be "nuts" to do whatever you suppose isn't/is "nuts"? -- Fullstop (talk) 20:29, 16 January 2009 (UTC)
  - take your head out of your arse for a minute and switch your humor module on. Semitransgenic (talk) 12:51, 17 January 2009 (UTC)

Physician, heal thyself --Philcha (talk) 12:56, 17 January 2009 (UTC)

Keep Slrubenstein | Talk 20:53, 16 January 2009 (UTC)
Comment.: Note to new additions: A sustainable argument for keep has not once, ever, been put forward. Never. So, if you think you have a sustainable argument for keep, make it now, and that means more than just "I think so". With no sustainable arguments for keep, and no sustainable arguments against not-keep, there is only one opinion (with any weight) that is evident. See also WP:VOTE on guidance. -- Fullstop (talk) 22:29, 16 January 2009 (UTC)

"With no sustainable arguments for keep, and no sustainable arguments against not-keep, there is only one opinion (with any weight) that is evident." Only if the information in dispute is relatively new, which is not the case here. --Ronz (talk) 23:58, 16 January 2009 (UTC)

Fullstop, why are you pointing us to WP:VOTE after you yourself have voted? And your analysis of the argument so far is merely your opinion so please don't threaten editors to "comment or else..." I think it makes sense to divorce the PSTS definitions from policy, but don't know where on WP they would go instead. I'm happy for policy pages to choose to refer to PSTS or not depending on WP opinion of editing policy. What I'm not happy is for WP opinion on editing policy to influence PSTS definitions. Colin°^Talk 00:10, 17 January 2009 (UTC)

"What I'm not happy is for WP opinion on editing policy to influence PSTS definitions" thoroughly baffles me. As best I can tell there is essentially full agreement with NOR and therefore isn't any "opinion on editing policy" that could influence the definition or does influence it. Good faith attempts are being made (not by me) to try to find PSTS language that does not cause concern. There has long been discussion centering on that, but it's far too much of a stretch to call that "opinion on editing policy influencing PSTS definitions." Were the PSTS definitions solid enough there'd be no discussion. The flaw is in PSTS itself. Minasbeede (talk) 00:18, 17 January 2009 (UTC)

Uncertain: I fully agree with the content of the section, but some parts of it have little to do with original research; in particular, the distinction between tertiary and secondary sources is completely irrelevant to original research. What's interesting about this section is that it has very little force of policy, because it doesn't actually forbid anything that isn't already forbidden by NOR; it doesn't restrict the use of tertiary or primary sources. It just provides broad suggestions. The last sentence reinforces that it's not to be taken as restriction. In light of the fact that this section is already written in the form of a guideline, it might make sense to either move it to a guideline page, or simply mark this section as non-normative guideline material. Dcoetzee 23:54, 16 January 2009 (UTC)
Keep. In the areas I edit in it is a very useful. --PBS (talk) 01:08, 17 January 2009 (UTC)
The three areas I tend to work in are military history, international law, and subjects with a very high POV content like genocides in history. If people were allowed to string together primary sources in those types of areas, then restrictions on OR would go out of the window. While I appreciate the arguments for moving this section out of NOR into a guideline, it would make the task of stopping the incorrect use of primary sources harder. How many seconds would it be until someone quoted "Policies are considered a standard that all editors should follow, whereas guidelines are more advisory in nature." ... ? Neither should it be deleted. I also appreciate that in some areas of interests it is much harder to decide whether something is or is not a primary source, but once demarcation has been made the principles are the same (where this line lies can either be decide in project groups or on the local talk pages or both). --PBS (talk) 11:04, 17 January 2009 (UTC)

Wait a minute. Not singling out primary sources as being particularly prone to misuse is not the same thing as allowing people "to string together primary sources in those types of areas." If the stringing together amounts to OR then it is already forbidden by WP:NOR and the type of source used doesn't matter: forbidden is forbidden, period. Let me be very clear: I do not believe PSTS is needed but I am not in favor of allowing people "to string together primary sources in those types of areas" in any manner that constitutes OR. [I italicized "allowing" because the use of that word seems to imply that there are editors and there are uber-editors, with the uber-editors having the privilege of deciding what ordinary editors may do. (If this is a bad characterization please correct it - I don't want a bad characterization to interfere with discussion.) I don't think that is the nature of Wikipedia.]

It is not forbidden to string together secondary sources. It is only this section that forbids the stringing together of primary sources. -PBS (talk) 19:58, 18 January 2009 (UTC)

If some editors are prone to such stringing together" then it is inevitable that those editors will see their edits reverted. When that happens they could be pointed to a PSTS guideline article that is designed to better explain why they should not do that kind of editing. I have no illusion that such a guideline will ever be complete, but it surely could approach that. Note particularly that I do not see removing PSTS from WP:NOR as in any way making the feared type of OR acceptable nor do I see the removal of PSTS in any way impeding the reversion of edits that constitute OR.

You have not addressed the problem of the difference between policy and guideline "Policies are considered a standard that all editors should follow, whereas guidelines are more advisory in nature." --PBS (talk) 19:58, 18 January 2009 (UTC)

You appear to have a concern that not having PSTS in WP:NOR invites abuse. I have a concern that having PSTS in WP:NOR invites a different kind of abuse: wikilawyering or wiki-whatever-the-word-would-be-for-brutal-reversions. My response to your concern is that if an edit constitutes OR then that edit may/must be altered or removed so that the OR is not part of Wikipedia by WP:NOR as it was before any source typing ever crept in. My concern seems to never really be addressed. (If an edit deserves a reversion then so be it. a "brutal" reversion would be one that amounts to "you used a primary source - that's a non-no," with no further guidance.) This is the real world: imperfections will exist. We cannot, by altering the wording in policies, make the world perfect (nor even the tiny corner of the world known as Wikipedia.) My view is that PSTS in itself is an imperfection. As such it can easily be dealt with: make it a guideline and remove it from WP:NOR. To my mind that in no way permits any OR that PSTS is claimed to forbid. PSTS would still exist as an attempt to provide guidance to editors who appear to have misused a source. Those editors who would cite PSTS as a reason to revert can remain as alert as they are now to such abuses. All I'd have them change would be to cite WP:NOR as the reason for reversion rather than cite the PSTS part of WP:NOR as the reason. They could still point the editor affected to the PSTS guideline. If PSTS is a guideline then much of the pressure (as evidenced by the years of discussion) to make PSTS actually say the right thing would go away. That would (I hope) mean this discussion would end, to the benefit of all. Minasbeede (talk)

The complaints of wikilawyering cut both ways unless the person making the accusations are making it because the want to ignore Wikipedia policy. As I explained above there is a difference between summarising secondary sources and primary sources in (#A new approach to the PSTS question). This is the only section of the OR guideline which highlights this difference. --PBS (talk) 19:58, 18 January 2009 (UTC)

Delete the material on secondary and tertiary sources (see proposal below). --Phenylalanine (talk) 01:12, 17 January 2009 (UTC)
- Keep. When available secondary sources are necessary to build a good article. That is worth mentioning in the policy. --Phenylalanine (talk) 11:57, 20 January 2009 (UTC)
Keep Central concept, long held by the community. Strawpolls never trump established policy, BTW. FeloniousMonk (talk) 03:09, 17 January 2009 (UTC)

Sure, of course straw polls never trump established policy. I'm trying to get a handle on what the extent of objection is here, in light of some lengthy reasoned criticisms and in light of several assertions on this talk page that consensus has not been demonstrated for WP:PSTS as a core policy. I count about five or six users who've held this basic position on this talk page. If there are more, we should find out and bring it more directly in front of the broader community. It's turned out to be very high maintenance; and the opposition, though scattered and limited to a few users, has been quite vocal and at length. If PSTS's opponents represent a broader constituency, by all means let's begin to find out, and if not, then it's time to find that out too and stop responding to a lot of the extremely lengthy criticisms of it. Either way is OK with me, but I've only got limited time to devote to apparently endless arguments. ... Kenosis (talk) 03:58, 17 January 2009 (UTC)

I'm not sure how these polls work, but if you ask people out of the blue if they think some large chunk of the policy should be removed, I'm not surprised if many would be highly skeptical. Wouldn't it be preferable to clearly lay out the arguments for and against such a change so that editors have some sense of why they are voting in the first place? --Phenylalanine (talk) 04:12, 17 January 2009 (UTC)

It's just a straw poll. The assertion has been made that WP:PSTS doesn't belong as policy and/or that its development was haphazard and never really achieved community consensus as a policy. If the opposition to WP:PSTS-as-policy is broad and not limited to just a relatively few vocal opponents, a poll ought give some preliminary indication of the possibility that a lot of WP users oppose WP:PSTS as policy. Perhaps it should instead be a guideline on its own page; or perhaps it has no place in WP at all; or perhaps it should not have the language limiting use of primary sources to supporting only descriptive content of what's in a source that can be double-checked without expert knowledge, thereby requiring a secondary source in all topics that require specialized knowledge (if this is removed, then there's no need for WP:PSTS as policy and it's properly reduced to a guideline without any real change in its effect, because the only real policy statement is essentially that we're limited to expressing in writing what's actually in the source, be it primary, secondary or tertiary, thereby rendering the PSTS distinction pretty-much completely moot, at least as a policy). Whatever the grounds, this is only a straw poll to try to discern who really objects. Because if there are not a lot of objections to WP:PSTS, then probably the proposal to remove it from WP:NOR goes nowhere, as FM just indicated somewhat to my chagrin. So, if I were a strong opponent of WP:PSTS in its present form, I imagine I would want to to leave a talk-page note or send an email to everybody I know who opposes it, including those who've opposed it on this talk page before, because here's a chance to make a case that there's this major, broad opposition to this WP:PSTS policy. Like I said, it's just a straw poll to get a very rough idea of whether the lengthy complaints have really achieved a lot of supporters that might call into question the notion WP:PSTS is a reasonable and sustainable part of WP content policy, but rather is arguably un-reasonable and un-sustainable and hence might belong somewhere else other than on a policy page. ... Kenosis (talk) 05:49, 17 January 2009 (UTC)

The question is not whether it is central or not, the question is where the PSTS section belongs. Discussion of primary sources is directly relevant and helpful, not so for secondary and tertiary sources. The current PSTS section belongs at WP:NPOV or in on separate guideline page, not here. --Phenylalanine (talk) 03:35, 17 January 2009 (UTC)

The only real question in this admittedly limited "straw poll" is whether one advocates that WP:PSTS should be kept in WP:NOR or should it instead be deleted/removed from its policy status in WP:NOR. I should think it's a reasonable opportunity for all opponents of WP:PSTS-as-policy-in-its-existing-form to get together under one tent for a change. That tent can reasonably be described as follows: If you think that "WP:PSTS in its existing form doesn't belong here in WP:NOR", then please say "delete". Either way, please feel free to articulate your perspective at reasonable length, and if need be to continue such argument(s) at greater length in a separate section. ... Kenosis (talk) 06:30, 17 January 2009 (UTC)

Keep as a significant and useful part in showing and determining the line between original research using the subject of the article as a source, and finding expert opinion on the subject which is to be accurately summarised, as well as providing the required third party basis of articles and sections of articles. Recent examples include a perennial debate at Talk:Expelled: No Intelligence Allowed where numerous editors want to disregard expert opinion, and base the article on their own impressions of the film. From another aspect, there's a wealth of primary material available on Charles Darwin and I'm careful to work within what expert historians say about this material, relying on these secondary sources conclude about the material rather than summarising the primary source without such references in a way that would inevitably mean me drawing my own conclusions. Primary sources are invaluable as part of the process, but the care set out in the PSTS section is needed to avoid original research. . dave souza, talk 11:30, 17 January 2009 (UTC)

This troubles me. You equate secondary sources with "expert opinion" and then base your reasoning on that. Surely there are secondary sources that are written with a strong bias (such as would happen with almost everything written in the old USSR. Milder forms of that sort of bias exist today.) Such secondary sources should normally be shunned by Wikipedia editors (shouldn't they?) So what you really mean is "secondary sources that are also relatively unbiased expert sources." Which means care is needed. Which is no different from care being needed in the use of primary sources. If proponents of PSTS always pick examples in which the primary source chosen is used badly and the secondary source chosen is used well all they're really doing is finding examples of lack of care for primary sources and the application of care for secondary sources. It isn't the source type that is inherently bad or good, it's the style of the editor who didn't edit with proper care when the primary source was used. There are no doubt things in Darwin's writings that could be quoted or referenced in an article about Darwin. What really matters is if the quoted material enhances the article. A priori it is probable that in general the secondary sources will be more condensed or broader and thereby more suited for a Wikipedia article. That's what matters, that's what promotes secondary sources over primary sources with regard to Darwin. If, however, a particular quote or reference highlights Darwin's early appreciation of something that only later was explained that could enhance the article by showing that Darwin recognized some particular kind of effect was occurring even though he had no glimmer of how that worked. If that were appropriate for the article in which it was quoted it would be a good use. But if all I'm doing is describing a particular kind of care in the use of primary source material then I'm doing no more than illustrating the need for care, and such illustration is secondary to the purpose and thrust of a policy. Minasbeede (talk) 14:46, 17 January 2009 (UTC)

Minasbeede, there is no conflict between the claim that view is "expert" and the claim that a view is "biased." Science is predicated on argument and debate. One of the flagship journals of anthropology, Current Anthropology has a "comment" section after all major articles where a dozen or so scholars have a chance to provide alternative or even opposing views. This is a good thing among experts. Sure, one hopes that experts eventually reach some consensus, but the process towards consensus can take years and involve contentious debates that are often quite productive. Dave Souza is making a good point about NOR - expert views should go in an encyclopedia, and given how contentious the sciences are, one criteria we can all agree on is that thy should at least come from verifiable, reliable sources and peer-reviewed journal articles (secondary sources) are a good example. You are making a point about NPOV: we should provide all significant views found in notable reliable sources. I agree completely. If there are diverse viws among physicists concerning string theory, or diverse views among anthropologists concerning the classification of a particular hominid fossil, these various views should go in the article. But the point about peer-reviewed articles, or other standards for "reliability," are indeed important because the views we include really should use care with data. And Wikipedia editors are in no position to assess the care with which one of us uses primary sources, and I have seen too many edit wars where one person thought he was using primary sources with care and many others thought he was completely twisting the data. NOR helps us avoid those conflicts AND maintain the quality of articles. And it does not conflict with NPOV Slrubenstein | Talk 16:35, 17 January 2009 (UTC)

As far as I can tell I intended absolutely no point about NPOV. My point was that some secondary sources are simply not suitable for use in Wikipedia (with the secondary point that some uses of secondary sources are inappropriate as well.) I can appreciate that in science there is argument and debate. Taking your example, is it probable that someone could quote something from the comments section after an article in Current Anthropology and have it be suitable, given that, as you say, it may consist of "alternative or even opposing views"? Perhaps that would sometimes be appropriate for giving due weight to opposing views but my question is directed more at taking something from the comments section and making it part of the article. The issue before us isn't how science is conducted, the issue is what is suitable material for an encyclopedia. Unless it can be shown that it is only with primary sources that some editors erroneously believe they are using the source with care when they are not the example you give illustrates only that care is needed, which I assert and which seems to be the sufficient necessary policy. Disputes will surely arise, errors will surely occur, but does singling out primary sources in a rather vague way actually assist any editor in deciding how to edit? I don't dispute NOR, I dispute PSTS. The positive glow from NOR does not transfer to PSTS. You say "Wikipedia editors are in no position to assess the care with which one of us uses primary sources." I don't think that is true but if it is how does PSTS as now worded or as ever worded or as it might some day in the future be worded turn the impossibility you claim into a possibility? One might cite the eating of draft animals with gusto by Antarctic explorers as evidence of their straits but who would not perceive that such a citation does not really indicate that in the general case draft animals are tasty or should be regular dietary fare? If an editor nonetheless made that claim how is PSTS needed for another editor to (very properly) remove that claim from wherever it appears? That claim would not be so much OR as it would be unjustified generalization. It isn't policy that cleans Wikipedia articles, it is the efforts by the body of editors. If PSTS is the justification for removal and is used without explanation that seems close to failure to grant the errant editor the assumption of good faith, with good faith requiring that an explanation be given - an explanation that does not simply invoke a policy section that has consistently been disputed. Minasbeede (talk) 16:59, 17 January 2009 (UTC)

I just realized that, rather astoundingly, you cite the policy/practice of a primary source (Current Anthropology) as justification for a policy that is intended to limit the use of primary sources in Wikipedia. Minasbeede (talk) 17:20, 17 January 2009 (UTC)

Current Anthropology is a peer reviewed journal that contains articles that are secondary sources by this policy's definition. Moreover, i still do not understand your point, which is about NPOV. NOR never says that any and all secondary sources must be included. The NOR policy is about how different kinds of sources should or should not be used'. It is NPOV that dictates what sources must and must be used. How vs. What, get it? And NPOV is clear: any significant view from a notable verifiable and reliable source must be included. Fringe views and views from unverifiable and unreliable sources cannot be used. Your point is already covered in NPOV. Slrubenstein | Talk 18:50, 17 January 2009 (UTC)

Nowhere in NOR does it say peer-reviewed journals are secondary sources. In the NOR section on RS it says: "In general the most reliable sources are peer-reviewed journals..." but that doesn't make them secondary sources. I am surely aware that there is vigorous discussion on this talk page about peer-reviewed journals and much of that discussion is necessitated by peer-reviewed journals being primary sources. As PSTS deprecates primary sources many feel a need to exempt peer-reviewed journals from that deprecation. My point isn't in any way connected to the notion that "any and all secondary sources must be included," it is that secondary sources, too, need to be used with care. If I assume that peer-reviewed journals are equivalent to secondary sources I do not see that in any ware removes the need for care. My emphasis is on care, which needs to be exercised with all types of source. From that I conclude that no special mention need be made of primary sources since the special mention states that care need be used, which is not unique for primary sources. Perhaps someone could make a count of all the sources that need to be used with care and find that a higher percentage of primary sources require care but that's not information useful for making policy. If care needs to always be used then that's fully adequate for a policy statement.

This is from the "Manuscript Submission" section on the Current Anthropology web site: "Articles should be at the forefront of present-day scholarship and may take controversial positions; they may even be speculative at times, as long as their intention is to bring important questions into scholarly debate." Is that the standard for inclusion in Wikipedia as well? Minasbeede (talk) 19:33, 17 January 2009 (UTC)

It is the standard for inclusion in CA. Do you really not understand what the English words "Manuscript Submission" means? Please get a dictionary! Slrubenstein | Talk 15:07, 18 January 2009 (UTC)

Not all primary sources are peer reviewed. Not all peer reviewed sources are primary sources. Even if it should be the community's will to remove the PSTS clause from this policy, it will only help the decision to write a clearer description of what are primary et all sources, in line with Colin's proposal, because the confused definitions are unavoidably compounding the confusion about what's appropriate in NOR policy. Professor marginalia (talk) 19:44, 17 January 2009 (UTC)

No field has precise, hard and fast dividing lines between primary, secondary and tertiary anyway, so it doesn't matter all that much. It's a tar pit-- unless one acknowledges that local consensus process is generally able to sort it through with discussion and adequate research to know how the information flows within a given field. If one acknowledges this, the general sketches and examples presently given in the policy are more than adequate to serve the purpose of the policy. The rest is, frankly, just picking nits. Same as with WP:WEIGHT as part of WP:NPOV-- you can't define in advance the exact parameters that constitute weight. Same as with WP:V#Reliable_sources -- one can't define in advance, from a policy page that overarches the entire wiki, exactly what the RSs will be in a given topic; the assessment of "reliability" quite commonly needs to be adjusted based on the evidence that's available to editors working in a particular topic area. ... Kenosis (talk) 22:57, 17 January 2009 (UTC)

Keep' As per Slrubenstein, Dave Souza, and PBS. dougweller (talk) 16:46, 17 January 2009 (UTC)
Remove. Would further point out that a section of a policy that is getting this high a level of opposition does not have consensus. (i.e. that the burden of proof here is on those saying this is consensus for the policy) Phil Sandifer (talk) 16:41, 18 January 2009 (UTC)
Keep. This is a crucial piece of the NOR policy, which assists in keeping out much Original Research. I understand the desire of those who wish to use Wikipedia to document recent popular culture to do away with the stricture, but allowing an exception for Original Research in this way would damage Wikipedia far more than any appreciable benefit. Wikipedia's issue is with its reliability, not with its comprehensiveness on topics related to recent popular culture. And regarding Phil's statement, I would counter that five people disputing longstanding policy does not mean the policy does not have consensus. The "burden of proof" is always on those wishing to change policy. Jayjg ^(talk) 20:41, 18 January 2009 (UTC)
Keep. Per Jayjg. And because it forces editors to think about the different types of sources and how to use them. If editors don't know the difference between a primary and a secondary source, they will invariably add their own analyses to articles. That's not a great problem in Ratchet & Clank, but it clearly is a big problem in any field where there is an established body of secondary literature. Jayen 466 22:22, 18 January 2009 (UTC)
Keep per Dave Souza & Jayjg. Jakew (talk) 22:23, 18 January 2009 (UTC)
Keep. This is an important part of the OR concept, which often helps to resolve disputes by making sure we defer to authoritative secondary sources over primary sources in cases of conflict. SlimVirgin ^{talk|contribs} 00:52, 19 January 2009 (UTC)
- The current policy does not say that at all, either explicitly or implicitly... — Carl (CBM · talk) 01:25, 19 January 2009 (UTC)
Keep per Jayjg. -- Vision Thing -- 21:14, 19 January 2009 (UTC)
Keep PSTS in NOR. We need more education and support around this structure, but the underlying concept is an important one. For those that think otherwise... I've got an editor at Hyperhidrosis (excessive sweating) trying to tell me that a single surgical experiment done to five dogs (which do not sweat) in 1973 "proves" that all humans who have a somewhat similar surgery done specifically to reduce sweating will have exactly the same changes to their heart rates (in apparent defiance of generally accepted medical practice). It's unbelievably easy to piece together stray bits of the medical literature to reach any conclusion you wish, if only you rely on case studies, what happens in a test tube, and so forth. WhatamIdoing (talk) 05:53, 20 January 2009 (UTC)
Delete, given that the section does not have consensus, is unclear, is not based on longstanding academic principles, and is not followed anyway. COGDEN 00:48, 23 January 2009 (UTC)
Keep. A crucial component of this policy. Crum375 (talk) 02:17, 23 January 2009 (UTC)
keep I've been keeping some track of the discussions on this topic although not commenting due to real life busyness but I'm just not convinced by any of the arguments for removal. JoshuaZ (talk) 02:54, 23 January 2009 (UTC)
Keep Although I doubt anyone will allow a fundamental change in this core policy based on what we say here, there isn't a conceivable structure for an NPOV encyclopedia that does not include PSTS. This is a ridiculous. OrangeMarlin ^{Talk• Contributions} 03:56, 23 January 2009 (UTC)
Keep Over at the Conservapedia article we pretty much depend on PSTS to stop the article descending into a pile of everyone's favourite snippets from CP. I'm also using it to try to stop someone using an autobiography as the main source in a biography article (the argument is here). Totnesmartin (talk) 09:59, 24 January 2009 (UTC)

Proposed addition

To the subsection, "Using Sources," in the "Sources" section, I would like to propose adding this line: "Remember that citations must accurately reflect the context of the source, and take care to avoid the original research of introducing a meaning not intended in the original source." Credit for the wording and suggestion go to Dave Souza, who suggested this on another talk page. I think it is a good idea. The point is that a quote may be taken out of context; a quote is a sample from a source, and should be used to make the point the source makes (doing otherwise would be original research). The suggested addition is meant to clarify and reinforce this point. Slrubenstein | Talk 19:04, 16 January 2009 (UTC)

The start of section "Sources" already says, "Research that consists of collecting and organizing material from existing sources within the provisions of this and other content policies is encouraged: this is "source-based research", and it is fundamental to writing an encyclopedia. Take care, however, not to go beyond what is expressed in the sources or to use them in ways inconsistent with the intent of the source, such as using material out of context. In short, stick to the sources" (emph as in the original) --Philcha (talk) 19:37, 16 January 2009 (UTC)

The suggested addition is meant to clarify and reinforce this point. Slrubenstein | Talk 20:28, 16 January 2009 (UTC)

I agree with this change - it's not clearly implied by the current text and is important. I would reword slightly to "take care to avoid introducing a meaning" - no reason to emphasize that this is original research, the whole page is about OR. Dcoetzee 23:57, 16 January 2009 (UTC)

I agree with proposed addition. It seems like a useful reinforcement. -- Vision Thing -- 21:18, 19 January 2009 (UTC)

I wouldn't mind seeing this point briefly reiterated in the short subsection on WP:NOR#Using_sources. Though it should be obvious that the expectation is that we not introduce any meanings that are not clearly intended in a given source, occasionally some users seem to miss this very basic expectation about sourcing. FWIW, I imagine that on some occasion(s) I've deviated from this expectation myself, e.g., when hastily seeking a source for something I've added to an article. A statement such as proposed by Dave Souza and Slrubenstein would concisely make this expectation explicit in the context of "using sources". ... Kenosis (talk) 17:45, 22 January 2009 (UTC)

Support, per above. --Philcha (talk) 18:14, 22 January 2009 (UTC)

Discussion of interest regarding WP:NOR

A discussion is currently ongoing at the talk page for WP:NOT suggesting that Wikipedia be redefined as journalism. I feel doing so goes against WP:NOR and all it stands for. The discussion in question can be found here. 23skidoo (talk) 19:28, 18 January 2009 (UTC)

Do WP:NC and WP:NOR conflict?

I understand, appreciate and strongly support the spirit and intent of WP:NOR as it applies to article content, but I'm having trouble seeing how it applies to article naming, particularly with respect to the processes of determining the most commonly used names (e.g., using the google test), the names most easily recognized name, and creating new names that do not conflict with other uses of the same name, all of which entail "original research" to some extent by their very nature. I've started a discussion on this at WP:NC, including a proposal suggesting that WP:NC explicitly state that WP:NOR does not and cannot apply to the process of article naming. Something to that effect probably should be stated here too. Comments? (please read my post there first - thanks). --Born2cycle (talk) 20:52, 30 January 2009 (UTC)

I don't regard the title of an article as part of the encyclopedic content of the article. Thus I don't believe that the NOR policy applies to it. Certainly we should refer to published sources in determining which name to use, but it will often be a matter of case-by-case judgment in the end.

It's also worth keeping in mind that the "no original research" policy does not prohibit editors from conducting research in the course of writing an article. The NOR policy only prevents editors from presenting their own conclusions in the article text itself, unless those conclusions can in principle be sourced. — Carl (CBM · talk) 22:34, 30 January 2009 (UTC)

I agree that most of the time, NOR would not apply to article names... however, I have frequently seen WP:NOR cited in naming disputes connected to articles on various WP:FRINGE theories. Such debates usually happen when an article is named in a way that promotes the Fringe theory. The argument usually centers on the argument that the article name itself constitues original research... that no one has ever called the Fringe theory by the name used in the article. What I am saying is ... I think WP:NOR can be an issue in article naming, in certain rare situations. It is often a case by case determination, as Carl states. Blueboar (talk) 22:45, 30 January 2009 (UTC)

Hi Carl and Blueboar. I understand and appreciate what Carl says in the second paragraph.

As to Carl's first paragraph, that's exactly my point, and your help and support in getting that notion effectively conveyed (including incorporating Blueboar's caveat) at WT:NC (and eventually WP:NC) would be appreciated. Currently, at WT:NC (flora), for example, editors are justifying use of certain verifiable/sourced names (e.g., the scientific Latin name for plants) for article titles instead of determining the most commonly used common English name essentially because determining the most common name is violating WP:NOR, is not verifiable, etc. What many of them seem to be seeking is to avoid precisely the case-by-case judgment that both of you note (I don't recall encountering either of you before, but your WP experience is obviously broad and deep), and that WP naming, because use the most easily recognizable name is inherently subjective, requires.

Now, whether using scientific Latin names for all, or almost all ("Scientific names are to be used as page titles in all cases except..."), plant articles, rather than common names, is a worthy idea is a separate issue (addressed here, for those who may be interested). My concern here is with the reference to WP:NOR at WP:NC, and the implication (at least to some) that it applies to the process of determining article names, disambiguating, etc. So, again, help on that issue, at least by posting here, would be appreciated. --Born2cycle (talk) 23:09, 30 January 2009 (UTC)

Determing whether to use the latin or common English name as the title on flora articles is not within the scope of WP:NOR ... as neither name would qualify as original research by our definition (unless the editor made up a name for the plant). Sorry, but I don't see any need to add anything about this to the NOR policy. Blueboar (talk) 23:16, 30 January 2009 (UTC)

I'm not asking to add anything about flora naming to the NOR policy! I wrote, "whether using scientific Latin names... rather than common names ... is a separate issue", and provided a link to the flora talk page for that discussion for those who may be interested (apparently not you! ;-) ).

I am suggesting that something be added to the NOR policy (and to the NC policy) to clarify that it does not apply to the article naming process, except for the caveat situation you mentioned, at least with respect to determining the name most easily recognized by English users, which is arguably a process that inherently requires original research. What do you think of that? --Born2cycle (talk) 23:30, 30 January 2009 (UTC)

I think it is instruction creep and really is not needed. In either policy. Blueboar (talk) 04:31, 31 January 2009 (UTC)

Then how do you suggest we deal with editors who argue that "use the most recognizeable name" and "use the most common name" (from WP:NC) should essentially not be followed because doing so is a violation of WP:NOR? I mean, technically, their argument is correct. --Born2cycle (talk) 02:54, 3 February 2009 (UTC)

See my essay User:Dcoetzee/Named topic bias. I argue in effect that more article titles should be OR. Dcoetzee 03:37, 31 January 2009 (UTC)

I agree with Blueboar on this. The storm in an organic cup of tea will blow over. --PBS (talk) 17:31, 31 January 2009 (UTC)

Then same question to you that I ask of Blueboard above. And flora is not the only place this argument has been made, and will continued to be made. Only a clarification will clear it up. --Born2cycle (talk) 02:54, 3 February 2009 (UTC)

Help!

Changes have just been made at WP:NC and WP:Naming_conventions_(common_names) that I believe wrongly give precedence to the jargon of specialists over that of non-specialists (assuming there is a conflict) in the process of choosing a name for a WP article title. I've started a strawpoll to see if there really is consensus for this. Your participation would be greatly appreciated. --Born2cycle (talk) 04:22, 3 February 2009 (UTC)

Mathematical OR

It would be nice to have a statement as to how far logical or mathematical derivations from a source are admissible and do not fall under OR.

I guess, we all agree that the following would not be OR. Source claims that there are 6,543,210,000 people on earth. Editor writes there are more than 6.5 billion people on earth[ref].

However, other logical derivations are much more complex and not obvious to all editors, such that their brandishing as OR or requesting a source might make sense.

I came here to find a policy on this problem or at least a guideline, but did not succeed. I would appreciate, if someone gave me a hint where to find this or, if necessary, starts a discussion on how we could formulate something reasonable. Tomeasy_{T C} 18:49, 4 February 2009 (UTC)

There's been a lot of discussion of this, and sooner or later we ought to codify it in a guideline or at least an essay. I'll let someone else rehash the consensus position. Dcoetzee 19:42, 4 February 2009 (UTC)

The policy states that, "The "No original research" rule does not forbid routine calculations (e.g. adding or subtracting numbers, rounding them, calculating percentages, converting them into similar units, putting them on a graph, or calculating a person's age) that add no new information to what is already present in the cited sources." See Wikipedia:These are not original research#Obvious deductions for more details. --Phenylalanine (talk) 00:58, 5 February 2009 (UTC)

I imagine Tomeasy is referring to the discussion at Talk:Tetrahedron#Newly_discovered_property_of_Regular_Tetrahedron?. A mathematical proof should be verifiable to a reliable source. Non-routine substitutions by WP users of essential components of the proof or alterations of the form of mathematical proofs are not routine calculations, and would very arguably be original synthesis. If the proof is found in a solid, reliable source, it should be used as presented in that reliable source and cited to that source. Such a usage is not original research.
..... If I understand the discussion in Talk:Tetrahedron#Newly_discovered_property_of_Regular_Tetrahedron? correctly, there appears to be some question whether the assertion about the proof under discussion is reliable, particularly w.r.t. the types of tetrahedra to which it applies. This issue isn't a WP:NOR issue, but rather a WP:RS issue. Please also note that WP:FRINGE or WP:WEIGHT could quite conceivably apply to a newly published proof. In general any questions about the proof's extent of acceptance in the mainstream mathematics community should be accordingly noted if the proof is used in the article (e.g. with a statement like "X has proposed the following proof for Proposition Y"). ... Kenosis (talk) 04:42, 5 February 2009 (UTC)

COI rephrase

Ryan Paddy changed the text about COIs:

Original text

It should be noted, however, that editing articles related to or about yourself, by you or those closely related to you, is strongly discouraged, and you are instead encouraged to discuss potential edits to such articles in the relevant talk page.

New proposal

It should be noted, however, that editing articles related to yourself in order to promote your interests is strongly discouraged, and you are instead encouraged to discuss potential edits to such articles in the relevant talk page.

I have reverted it because it is less clear about COIs involving those related to the editor (e.g., friends, family, employer). Also, editing in order to promote your interests is not just strongly discouraged; it is not permitted. Editing articles related to yourself in ways that don't promote your interests is discouraged (because so many editors have trouble recognizing when they're promoting their interests). WhatamIdoing (talk) 04:01, 9 February 2009 (UTC)

I don't know when the COI thing was added, but it probably shouldn't be here, because this is policy and COI is a guideline. It's also not related to OR. It might be better just to provide a link. SlimVirgin ^{talk|contribs} 04:18, 9 February 2009 (UTC)

I made the change in order to make the discussion of COI in this article match more closely to what WP:COI actually says. It doesn't say that editing pages about yourself is strongly discouraged. That's an inaccurate summary, and I'd be interested to see a policy or guideline that actually says that. COI is really about inappropriate promoting of interests. I simplified the text to "articles related to yourself" because it was too wordy, and there is no need for such detailed clauses about closely related people here because they are covered in the WP:COI article, which is linked to. I still think my edit was preferable, unless a policy or guideline can be shown to say that "editing articles about yourself is strongly discouraged". If so, that article would be as relevant as WP:COI to link to here. Ryan Paddy (talk) 06:25, 9 February 2009 (UTC)

Does there need to be anything about COI in this policy, other than a link? SlimVirgin ^{talk|contribs} 19:01, 9 February 2009 (UTC)

The second paragraph in WP:COS re COI is a digression and should be deleted, in my opinion. --Bob K31416 (talk) 00:09, 10 February 2009 (UTC)

For reference, here's what the section looked like before the second paragraph was added on Sep 11, 2008 [40]

== Citing oneself ==

Shortcut

WP:COS

This policy does not prohibit editors with specialist knowledge from adding their knowledge to Wikipedia, but it does prohibit them from drawing on their personal knowledge without citing their sources. If an editor has published the results of his or her research in a reliable publication, the editor may cite that source while writing in the third person and complying with our neutrality policy. See also Wikipedia's guidelines on conflict of interest.

This appears to be the same as the present first paragraph except for the "See also..." at the end which seems OK and should be added. --Bob K31416 (talk) 00:47, 10 February 2009 (UTC)

Thanks for finding that, Bob. I've restored that version because COI is only loosely related to OR, and because we're in danger of elevating it to policy status without proper discussion. SlimVirgin ^{talk|contribs} 00:57, 10 February 2009 (UTC)

I agree with the restoration. --Bob K31416 (talk) 01:06, 10 February 2009 (UTC)

Good solution, there was no need to summarise COI here. Ryan Paddy (talk) 02:11, 10 February 2009 (UTC)

Strongly agree. Jayjg ^(talk) 05:59, 10 February 2009 (UTC)

Proposed change to WP:PSTS

Primary sources
Shortcuts
WP:PS
WP:PRIMARY

All analytic, synthetic, interpretive, explanatory, or evaluative claims about published material must be referenced to other reliable sources, rather than original analysis of the source material by Wikipedia editors.
Editors must be especially careful to avoid original analysis when using primary sources, which can be defined as providing "an inside view of a particular event".
For example, an account of a traffic accident written by a witness is a primary source of information about the accident. Other examples 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; original philosophical works; religious scripture; published notes of laboratory and field experiments or observations; experimental results written by the person(s) who conducted the experiments; and artistic and fictional works such as poems, scripts, screenplays, novels, motion pictures, videos, and television programs. The key point about a primary source is that it offers an insider's guide to an event, a period of history, a work of art, a political decision, and so on.^[1]
Primary sources that have been reliably published (for example, by a university press or mainstream newspaper) may be used in Wikipedia, but only with care, because it is especially easy to misuse them. Any interpretation of primary source material requires another reliable source for that interpretation. Without that other source, a primary source may be used only to make descriptive claims, the accuracy of which is verifiable by a reasonable, educated person without specialist knowledge. For example, an article about a novel may cite passages from the novel to describe the plot, but any interpretation of those passages needs another source.
Unsourced material obtained from a Wikipedian's personal experience, such as an unpublished eyewitness account, should not be added to articles. It would violate both this policy and Verifiability, and would cause Wikipedia to become a primary source for that material.

I propose we replace the PSTS section with something like this. Feel free to propose alternative wordings for the text, but I'm convinced that the focus should be on primary sources. We could maybe add a link to secondary source and tertiary source somewhere in the text, but more than that would be unnecessary in my opinion. --Phenylalanine (talk) 01:09, 17 January 2009 (UTC)

When I first read mention this proposal in the previous discussions, I thought it would be a good solution. Now that I have the well-written alternative above to compare with the current wording, I've changed my mind.

I think the current mention of "Wikipedia articles should rely mainly on published reliable secondary sources..." is important and extremely valuable to keep. As with the discussion of primary sources, this information helps editors resolve some of the most common OR/SYN problems.

I think the other information on secondary and tertiary sources is valuable, but doesn't fit into WP:OR other than to provide context for the rest. I certainly wouldn't want this information deleted, but moved to an existing or new policy/guideline. --Ronz (talk) 01:48, 17 January 2009 (UTC)

Ronz, "Wikipedia articles should rely mainly on published reliable secondary sources..." has everything to do with the Neutral point of view policy, but not much to do with this policy. When used with special care, primary sources pose no problem in regards to the original research policy. So explaining to people that they must be especially careful when using primary sources in this policy is helpful, but telling them to rely mainly on secondary sources in this policy is not helpful as it suggests that this is required by the original research policy, when it is in fact required by the neutral point of view policy. That is why the real focus is primary sources. --Phenylalanine (talk) 02:50, 17 January 2009 (UTC)

I agree that PSTS applies to NPOV and it deserves treatment there in addition to WP:OR, WP:RS, and WP:V. NPOV needs discussion of all three, primary, secondary, and tertiary; OR needs primary and secondary; and WP:V needs primary. The arguments for coverage in WP:RS appear to be mostly for context and completeness. --Ronz (talk) 17:09, 17 January 2009 (UTC)

The whole purpose of "should rely mainly on published reliable secondary sources" is to globally forbid using material from primary sources that indicates secondary sources are in error. Surely some of the time that is correct, surely some of the time it is incorrect. The guidance needed (and which can never be provided in a policy nor in a guideline) is how to tell the cases apart. Over and over it has been argued that sometimes (or the equivalent) this restriction is useful. "Sometimes" isn't good enough in a policy. Forget source typing, think only in terms of what policy is and isn't. A rule that is only valid part of the time is bad policy. Efforts that have been going on for over a year have not succeeded in solving the real problem, which is defining when "sometimes" is. How can the policy ever define, to the level necessary in a policy, when primary sources can and cannot be used? If the blow-this-issue-away answer is "it's up to the judgment of the editor" then sure, I agree. But if the judgment of the editor is the ultimate determiner then the presence of PSTS in the policy is entirely superfluous.

I hardly edit, this touches me slightly or not at all. Were I to edit I would try very hard to exclude OR from what I inserted and would in essence ignore PSTS. As I don't even know what PSTS really means (in terms of policy) ignoring it is the only option. Minasbeede (talk) 02:00, 17 January 2009 (UTC)

So, do you support my proposal? --Phenylalanine (talk) 03:09, 17 January 2009 (UTC)

If calling it WP:PS implies a separate guideline page, yes, I support your proposal - in particular, making this a separate guideline page. My memory may be faulty but I think your first paragraph pretty closely approximates what WP:NOR said back when it was only 500 or so words long. I think that was good. Minasbeede (talk) 03:17, 17 January 2009 (UTC)

I was actually thinking of having this text replace the PSTS section on the policy page. We could move the current PSTS section to a guideline page or the NPOV article. I'm surprised there is no discussion of PSTS at WP:NPOV. --Phenylalanine (talk) 03:28, 17 January 2009 (UTC)

Oppose Primary sources are not "those giving an insider view of a particular event." that's confusing them with COI, POV, or first person sources. Not that I think we have a clear definition, and I am not sure there is any useful one generally. In history, primary sources are the immediate contemporaneous records of events about which history gets written. This does not directly apply to any other subject. In literature, they are the works being discussed. In science, they're strictly speaking the usually unpublished laboratory records, and we misuse primary sources here to mean "primary journals", the ones that report such results. My own view on this policy page is that the terms should be deprecated on Wikipedia. —Preceding unsigned comment added by DGG (talk • contribs) 02:10, 17 January 2009

Hi DGG, I have no problem removing the definition and just having the examples illustrate what primary sources are. What do you think? --Phenylalanine (talk) 03:09, 17 January 2009 (UTC)

I think the starting point for any explanation is the way these categories are functionally related. Secondary sources are used as the principle sources for tertiary sources. Primary sources are used as the principle sources for secondary sources. "What" is a primary or secondary source can be fleshed out after this, but the main point is that we should think less about what something "is" and more about how it is "used." Slrubenstein | Talk 03:00, 17 January 2009 (UTC)

Agreed, which is why it is useful to explain that primary sources, as illustrated by the examples, are especially easy to misuse in relation to this policy. Secondary and tertiary sources are not really relevant to this policy. --Phenylalanine (talk) 03:18, 17 January 2009 (UTC)

I'm seeing neither a need to reword the original part of the policy nor an improvement in this version. I oppose the proposed changes. FeloniousMonk (talk) 03:11, 17 January 2009 (UTC)

The need is that secondary and tertiary sources are not relevant to this policy. The current PSTS section should be moved to a guideline page or to WP:NPOV. --Phenylalanine (talk) 03:18, 17 January 2009 (UTC)

I think the first para does the job clearly and thoroughly. A few more examples would be helpful, e.g. (assuming the citations stand up) "Authority A comments on Hamlet" is OK but "Authority A's comments on Hamlet are dubious" is not, while "Authority B wrote that Authority A's comments on Hamlet are dubious" is OK - etc., etc., etc.

Might also be helpful to say a little about the limtis of clear and obvious deductions, e.g. if a reliable census results table shows a population decline, it's OK to write "The population declined" (and that's all), but it would be improper for editors to express opinions on the development of Shakespeare's ideas or style based on output from computer analyses of his vocabulary and phrasing. --Philcha (talk) 12:07, 17 January 2009 (UTC)

Oppose any definition of primary sources that isn't (a) well sourced itself to a few good external sites (b) relatively comprehensive by listing several defining features (rather than the one we have a the moment which says merely "close to the event", or nonsense such as "insider's guide") (c) avoids examples of unpublished primary sources (there's no point in listing surveillance videos or laboratory notes, for example, as they are not the sort of primary sources that editors can actually cite) and (d) excludes sources that most authorities include (such as first publication of original experimental research) on the basis of dogma or effect on existing policy statements. Perhaps we should just link to the WP articles on primary and secondary sources. At least that way they have a half-decent chance of being reliably sourced, avoid original research, and might improve over time rather than being turned upside down before being gutted. (unwatching) Colin°^Talk 16:33, 17 January 2009 (UTC)

The source for the definition is in the footnote (the same footnote we have in the policy at present). Like I said, Im open to any wording changes. My proposal is just an example of what the section would look like if it focused mainly on primary sources, as opposed to the current PSTS section.--Phenylalanine (talk) 19:58, 17 January 2009 (UTC)

Oppose. By the way, the proposal uses a quote: "an inside view of a particular event" but you do not say who or what you are quoting. Where is this quote from? A citation please. Slrubenstein | Talk 18:53, 17 January 2009 (UTC)

The source for the definition is in the footnote. --Phenylalanine (talk) 19:57, 17 January 2009 (UTC)

Thanks. In that case, the note should immediately follow the quote. I happen not to find thse souces veryuseful. I think this is one case where Wikipedia can do better.

Say where you got it

Ok, here's a general question: Suppose I have access to all secondary sources available on a topic and I decide to write a wikipedia article on that topic from scratch based on those secondary sources, but instead of referencing the secondary sources in the 'pedia article, I decide to reference the primary sources that are used in these secondary sources. Is it possible for me to build a good Wikipedia article, without doing original research, by referencing only those reliable primary sources? --Phenylalanine (talk) 19:57, 17 January 2009 (UTC)

Are you asking about an article "based on secondary sources", but the only references given will be primary sources? Professor marginalia (talk) 20:05, 17 January 2009 (UTC)

Yes exactly. --Phenylalanine (talk) 20:07, 17 January 2009 (UTC)

Why do this? If I'm reading this right, it might even be improper practice-normally in academic work, when using content from a source the source should be credited. Professor marginalia (talk) 20:16, 17 January 2009 (UTC)

Apart from that, can I build a good 'pedia article, without doing original research? --Phenylalanine (talk) 20:28, 17 January 2009 (UTC)

I'm sorry to be so dense. If the claims in the article are "based on secondary sources", that isn't original research, whether you've cited those sources or not. But the 'pedia requires sources. Obviously, that means the sources have to be furnished. What am I not getting? Professor marginalia (talk) 20:37, 17 January 2009 (UTC)

If Professor marginalia is having difficulty with a hypothetical situation involving PSTS, what chance do less experienced editors have?

OTOH if you simply say "editors cannot use their own interpretations or evaluations of sources", it becomes fairly simple.

Example 1: If I have a bunch of census tables I can fairly say the population increased / decreased by X%. But if I want the WP article to contain conclusions about the social trends involved, I have to cite sources that express these conclusions, I can't just write in my own.
Example 2: I can quote a few lines from the "To be or not to be, ..." soliloquy in Hamlet. But if I want the WP artcile to include comments on the soliloquy's dramatic significance or use of language, I have to cite sources that state these conclusions.
If Authority A comments on "To be or not to be, ...", I can summarise Authority A's comments. But if I want to provide an evaluation of Authority A's analysis, I have to cite another source that provides such an evaluation.

This appproach avoids all the complications about "when is a secondary source not a secondary source?" Primaryness, secondaryness, etc. are not absolutes, they're relative to what is being written in the Wikipedia article. But explaining relative concepts is tricky - I'm aware of this because I work on paleontology articles, where cladograms are common. It's much simpler to ask, "Is what I am about to write a fair summary / representation of a citeable source?" --Philcha (talk) 22:09, 17 January 2009 (UTC)

I'm having trouble with the example given, not the policy. As it was described, it's a no-brainer. In the example given, you have secondary sources and you used them to write the report. So then? You cite your sources. That simple. This is the typical practice expected in research papers assigned in your average secondary school. It's not complicated. I fail to see how this illustrates there is a problem with the PSTS policy. Professor marginalia (talk) 22:45, 17 January 2009 (UTC)

In that case you need to get some sleep and come back in better shape tomorrow :-)

Doing it my way ("editors cannot use their own interpretations or evaluations of sources"), Phenylalanine's example is fairly simple. Using only the primary sources, editors are allowed only to summarise these. In most cases that would fail WP:N for the simple reason that it failed to prove that the rest of the world had taken any notice. In addition the article would not be satisfying, as it would be unable to explain the importance of the primary sources' content - unless they proclaim it themselves, which should probably be dismissed as self-promotion.

I'm taking the advice I gave you :-) Philcha (talk)

Philcha, maybe you do not have an ear for irony, but Professor marginalia is not being dense, he is being polite. I am pretty confident he understands the hypothetical cases, he just does not agree that they demonstrate a problem with the policy. He has explained why. Perhaps the problem is that you or Phenylalanine do not understand his/her reasons. Slrubenstein | Talk 23:40, 17 January 2009 (UTC)

Slrubenstein, did you not notice the smileys?

The real issue is not whether Professor marginalia has difficulties with PSTS, it's whether PSTS is the simplest and clearest way to achieve the objectives of WP:NOR. I have argued that it is not, because of the context-dependence of primaryness, secondaryness, etc., and because I think the simpler formulation "editors cannot use their own interpretations or evaluations of sources" does the job quite well enough, especially with the simple examples I gave above. If you think that is harder to understand than PSTS or leaves loop-holes whihc PSTS closes, please explain why. --Philcha (talk) 16:27, 18 January 2009 (UTC)

This is why. In my editing experience, NOR disputes are rarely over confusion about whether some text is or isn't a primary source. Disputants commonly deny the analysis or inference they're injecting to the article is merely their opinion from a reading of the material-they'll argue it's fact not opinion; they'll argue that it's "commonly known"; they'll argue it's a summation based on common sense not opinion; they'll argue the content is essential to provide the otherwise unrepresented "alternative view" to conform with NPOV, they'll sidestep the summary altogether and quote mine instead, waving the "let the reader decide" banner. OR they'll argue the primary source itself is the Real Thing and therefore more accurate to rely upon than the secondary sources that analyze or describe it. In these kinds of disputes, deferring to the best available secondary sources is critical. Professor marginalia (talk) 18:53, 18 January 2009 (UTC)

I don't think I'd need PSTS to dispose of the types of OR you've listed. In fact I think the complexities of PSTS would be a hindrance. I'd respond along the lines of "If that's not just your opinion, show me a good source that says the same thing." --Philcha (talk) 22:49, 18 January 2009 (UTC)

Not all of them in all their forms. But "show me a good source to verify this is noteworthy published opinion, not your opinion" is exactly what the PSTS policy instructs, doesn't it? The source for that confirmation of the opinion you've demanded will be, by definition, a secondary source, and it's silly to run away from calling it what it is in the policy. The conflict is often argued like this, "why do I need another source? it's what this text is about. the words speak for themselves. if you aren't knowledgable enough to understand what it means don't meddle. anyone knowledgable of this topic who isn't biased/stupid/blind would agree that's what it says/means/proves. what can another source prove? if it says the opposite, it's wrong. if it doesn't say, it's incomplete. another source can't be more accurate about what a text says than the original itself. others can be biased or ignorant. if I need one more source, why not two? three more? six more? what's the magic number? why don't you demand two sources for every claim in the article then? there's no call to single my edit out. so you're wrong, I'm right, go away." The PSTS policy helps explain how come. It helps explain why. Professor marginalia (talk) 01:42, 19 January 2009 (UTC)

I agree with the learned Professor. Jayen 466 02:58, 19 January 2009 (UTC)

(Undent) Phen, writing something based upon secondary sources, and then pretending that you based it on the primary sources that they cited does not technically violate WP:NOR (although how the casual inspector would know this is beyond me: perhaps you'd leave a note on the talk page to say that you lied about the sources you consulted?).

However, doing that directly violates Wikipedia:SAYWHEREYOUGOTIT, so it's prohibited by a different rule (not to mention normal academic standards for honesty). WhatamIdoing (talk) 06:00, 20 January 2009 (UTC)

I am aware of that rule and I acknowledge that, when available, secondary sources are necessary to build a good article. Let's move on shall we. --Phenylalanine (talk) 11:53, 20 January 2009 (UTC)

I want to point out that there are many times when we do want to cite the primary sources as well. Several are listed at Wikipedia:Scientific citation guidelines. Here's one relevant paragraph:

"If Wikipedia has an article about an eponymous topic – such as Michelson–Morley experiment, Sunyaev–Zel'dovich effect, Green–Schwarz mechanism, Alpher–Bethe–Gamow paper and Kaluza–Klein theory – then editors of this article should, if feasible, explain why the names are attached to the result or experiment. To this end, editors of these articles should consider citing the original papers, even if they are not used as sources in writing the article. However, articles that only link to an eponymous article might not cite the original papers, depending on context. In this case, a reader looking for a reference may easily click the article link to find it."

— Carl (CBM · talk)

CBM is correct-- primary sources are commonly necessary, for a wide variety of reasons across the wiki. ... Kenosis (talk) 05:49, 11 February 2009 (UTC).

Back to normal views

Oppose change. Has all the idiocy of the current version. Phil Sandifer (talk) 16:39, 18 January 2009 (UTC)

Who says what and the consensus

It seems to me the way Wikipedia is constructed Consensus can overrule any number of sources. Then there also is the risk of becoming a recycled media outlet. If we are honest both are original research. I don't have any suggestions how to improve this but it seems to be the main drawback towards writing quality articles. Drive by reverting has turned into a sport where the hard working contributors (right or wrong) are always the victim.

If we look at knol we can only conclude there are advantages in having named and validated contributors. Using your real name brings up additional trolling and off wiki stalking, it seems only fair to give such users something in return and allow experts in the field to write a good article. We don't want an article about a religion written by people who hate that religion. Things are more and more moving in this direction calling people "conspiracy theorists", "fringe" and having a "rational skeptics" police office.

Take for example practitioners of alternative medicine topics like Chiropractic, naturopathy, herbalism, traditional Chinese medicine, Unani, Ayurveda, meditation, yoga, biofeedback, hypnosis, homeopathy, acupuncture

Here it seems obvious to me we want those contributors to contribute as much good information as possible. A section written by untrained skeptics can be appropriate but this shouldn't expand into taking over the whole article and running off the professor or doctor.

A more vague example: I tried to add a paper to the cold fusion article, the paper was on newenergytimes, this seems to me to be a good website about the subject. This reference triggered the spam filter. One could call this website a lot of things but it's not spam. The website is just a mirror where the publication can be found. Here Wikipedia tends to turn into a biased media outlet where the actual researchers never get a chance to contribute to the actual article.

I feel it is safe to allow a bit more self citation if some one is an established expert in an area of expertise. Having children revert such contributions does not have a positive effect on the motivation. 10 unnamed/untrained editors are not more right than one expert. The expert may however have poor wiki editing skills.

Most journalists working for established news media are not professionals in the topics they write about, they write on a large number of subjects.

Wikipedia also has a responsibility to move forwards. If articles do not reflect progress people will think there is non. I imagine it can be incredibly humiliating to have your area of expertise described as a crank, fringe heap of quackery under the first search result by Anonymous users.

Again I'm not so much putting up a suggestion, this is more intended as a talking point to improve overall quality. I guess the main point is that the editor consensus does not reflect the source of the information nor does it show any expertise of any kind.

What are your thoughts?

84.104.135.86 (talk) 04:30, 11 February 2009 (UTC)

To clarify how I come to this suggestion:

http://en.wikipedia.org/wiki/Talk:Royal_Rife#Explanation_of_TS.27s_revert_of_an_edit_by_84.104.135.86

http://en.wikipedia.org/wiki/Talk:Free_energy_suppression#page_blanking_by_LeContexte_and_Oli_Filth

http://en.wikipedia.org/wiki/Talk:Oxyhydrogen#Big_deletions_by_OMCV

http://en.wikipedia.org/w/index.php?title=Free_energy_movement&diff=269875240&oldid=269870265

Permission to contribute was denied per wp:own? lol

84.104.135.86 (talk) 05:21, 11 February 2009 (UTC)

Hi. I think you mistakenly posted your messages in the wrong place. This discussion page is for discussing the No Original Research policy. I would suggest posting your messages at Wikipedia talk:Consensus and/or Wikipedia talk:Ownership of articles. Perhaps other editors may know of the appropriate noticeboards or places for you to go to, to look for help with the problems that you are experiencing. Links at Wikipedia:Questions might be of some help in directing you to the right place. Good luck. --Bob K31416 (talk) 05:55, 11 February 2009 (UTC)

Should we lock this page?

Hi. Should we? This is one of Wikipedia's law pages, and I feel it should be locked. Does anyone out there agree with me on this? 76.235.212.80 (talk) 22:08, 11 February 2009 (UTC)

No... the fact is, even after all this time, this policy is not perfect (none of our policies are). Editors need to be able to make changes in an effort to improve it. Of course, any significant changes that are posted without discussion and consensus will be immediately reverted, so the page is effectively locked. Blueboar (talk) 03:29, 12 February 2009 (UTC)

wp:syn is hard to understand

seriously. it should be clearer. the premise seems ok, but the example is funky at best. let's have it not be a policy unless it can be made a little clearer. Codigo 'll aka Huh? 02:46, 12 February 2009 (UTC)

How to apply WP:NOR's "Directly related" principle

This is an essay I wrote with the help of User:Jayen466. Any comments would be greatly appreciated. Thank you! --Phenylalanine (talk) 20:29, 1 February 2009 (UTC)

It looks like the essay is trying to show how to apply the principle of "directly related" to cases suspected of violating WP:SYN. The hypothetical examples that you give of providing context derived from reliable sources, appears to be directly related to the topic of the article. If that context is then used with info from another reliable source to make a novel conclusion that is not in any source, then that may be a violation of WP:SYN. As far as I can see, providing context without explicit synthesis is not a violation of WP:SYN.

However, one could object to the context if it is not the proper context. For example, in an article on fruit, there could be a section about apples. An editor might try to provide an inappropriate context for something in that section about apples, using a source that is about oranges. Then that orange context could be challenged because it is not directly related to the topic of apples, even though it is directly related to the topic of fruit. But it would depend on the specifics of the case.

--Bob K31416 (talk) 16:55, 7 February 2009 (UTC)

Thanks for your feedback Bob. As I read it, the policy does not make a distinction between explicit and implicit conclusions. Furthermore, on what grounds would explicit conclusions be forbidden, but not implicit conclusions. That would be an arbitrary rule in my opinion. Regarding the oranges example, if the information from the source on oranges provides useful context and is applicable to all fruits, including apples, I think it would be ok to use that source if none other are readily available. If there is agreement that the information is not helpful to the reader, of course it should be removed; and if the information is not applicable to apples, but is presented as such, we have another guiding principle which deals with exactly that: the source must "directly support the information as it is presented". Cheers, Phenylalanine (talk) 18:54, 7 February 2009 (UTC)

Re "As I read it, the policy does not make a distinction between explicit and implicit conclusions."

By "policy" I presume you are referring to WP:SYN. Perhaps that section doesn't make a distinction between explicit conclusions and implicit conclusions because it's not concerned with implicit conclusions. WP:SYN only discusses cases of A + B = C, not cases where just A and B are in the article without C. --Bob K31416 (talk) 20:48, 7 February 2009 (UTC)

Yes, NPOV is often an issue, but not always. The Smith & Jones example is a case in point. If you remove the conclusion (C), it can still be inferred from statements A & B with no false implications. On what grounds would explicit conclusions be forbidden, but not implicit conclusions? --Phenylalanine (talk) 21:13, 7 February 2009 (UTC)

Re "On what grounds would explicit conclusions be forbidden, but not implicit conclusions?"

On the grounds that implicit conclusions are not discussed by WP:SYN.

(P.S. Sorry about deleting part of my message just before you posted you message. I'd put it back but I edited it several times before I deleted it so I'm not sure which version you saw. ) --Bob K31416 (talk) 21:30, 7 February 2009 (UTC)

Since implicit conclusions are conclusions, since the WP:SYN policy does not expressly indicate that it applies only to explicit conclusions, and since interpreting the policy in such a way would defeat its purpose in practice, I think we ought to follow the spirit of the policy and treat implicit conclusions as we do explicit conclusions. --Phenylalanine (talk) 21:53, 7 February 2009 (UTC)

As far as I can tell, both the notion of implicit conclusion and the notion that it is in the spirit of WP:SYN that it be forbidden, come from you and not the Wikipedia. Sorry. --Bob K31416 (talk) 22:21, 7 February 2009 (UTC)

May I add that "directly related" does not seem to be the subject of your essay, but rather what you call implied conclusions from info, whether or not the info is "directly related" to the topic of the article. --Bob K31416 (talk) 23:03, 7 February 2009 (UTC)

A few more thoughts. Please note that a Wikipedia article tries to reflect what is in the literature regarding a topic and it may be that the facts in the literature suggest a conclusion when brought together in the Wikipedia. Such a conclusion may also be suggested if a person looked at the facts in the literature instead of looking at the Wikipedia. So to repeat, the Wikipedia is just supposed to reflect what is in the literature. If facts can be brought together that reflect the literature and they suggest a conclusion, then this is worthwhile, in my opinion, since that same conclusion could be suggested if the reader looked at the original literature instead of the Wikipedia.

However, a problem can occur when an editor selectively chooses facts from the literature, or takes facts out of context, to promote a view that violates WP:NPOV . It seems that this can be corrected without making WP:NOR more restirictive, which I think is the effect of the notion of "implicit conclusions".

One concern is that an editor could rampantly invoke the concept of "implicit conclusion" to exclude worthwhile material that simply doesn't suit the editor's personal preference. But also an editor may simply think he is conscientiously, albeit incorrectly, rooting out violations of WP:NOR using the proposed concept of "implicit conclusion" and in the process remove mostly worthwhile material.

Like I mentioned, it's choosing facts selectively or out of context to promote a position, that is the problem. This can be corrected by including the other facts and the rest of the context, or by pointing out that they were excluded so that the offensive editing can be deleted. --Bob K31416 (talk) 18:29, 9 February 2009 (UTC)

Information that is intended to suggest a conclusion in the literature may be presented as suggesting a conclusion within a Wikipedia article. In practice, this means that the information must come from the same source, and must be presented by the source author(s) in a way that clearly indicates a deliberate conclusion. Otherwise, an editor is engaging in original research by attempting to present information in an article in a way that serves to suggest a conclusion that was not deliberately suggested within the literature.

I agree with you that presenting information in a way that attempts to suggest a conclusion which does not strictly follow from the premises can be excluded on WP:NPOV grounds alone.[41] However, an implicit conclusion that properly follows from the premises may not violate WP:NPOV. The Smith & Jones case is an example where a neutral implicit conclusion is specifically forbidden on WP:SYN grounds.

In practice, editors will deal with the implicit synthesis rule in the same way that overt synthesis is handled. Just like overt conclusions, a clearly implicit conclusion does not depend on one's personal preference or state of mind; if it really is implied, any editor should be able to notice it. --Phenylalanine (talk) 01:31, 10 February 2009 (UTC)

I thought the link you mentioned [1] was pretty good and is something we both agree on. There are various things in your last message that I can comment on but I think I'll just try to get clarification on one, for now. It's looking like what you are concerned with is not false implications, but true implications, even if they are clearly true and based on facts that are directly related to the topic of the article and there is no specific conclusion made in the wiki. Is this one of the things that you are concerned with? --Bob K31416 (talk) 15:55, 10 February 2009 (UTC)

Yes, I am not concerned with false implications, but a specific conclusion has to be implied in the wiki that follows from a set of premises. Cheers, Phenylalanine (talk) 12:56, 11 February 2009 (UTC)

<outdent> Thank you. I'll continue now with another aspect of your essay.

Here's an excerpt from your essay re the Smith Jones example -

The policy identifies statements "B and C" as original synthesis. It says, "This entire paragraph [statements "B and C"] is original research, ...

(Just so there's no misunderstanding, I should clarify that Phenylalanine added the part in brackets, [statements "B and C"], to the WP:SYN quote.)

I think you misinterpreted what "entire" meant in the Smith Jones example in WP:SYN. Here it is for reference. I highlighted in bold the part containing "entire".

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

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

That much is fine. Now comes the unpublished synthesis of published material. The following material was added to that same Wikipedia article just after the above two sentences:

If Jones's claim that he consulted the original sources is false, this would be contrary to the practice recommended in the Harvard Writing with Sources manual, which requires citation of the source actually consulted. The Harvard manual does not call violating this rule "plagiarism". Instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them.

This entire paragraph is original research, because it expresses the editor's opinion that, given the Harvard manual's definition of plagiarism, Jones did not commit it. To make the paragraph consistent with this policy, a reliable source is needed that specifically comments on the Smith and Jones dispute and makes the same point about the Harvard manual and plagiarism. In other words, that precise analysis must have been published by a reliable source in relation to the topic before it can be published in Wikipedia by a contributor.

I think that what you misinterpreted in the above was the meaning of the word "entire", which is defined as "having no element or part left out : whole".[42] In so doing, you erroneously interpreted that the two sentences taken alone are original research according to the comments in WP:SYN regarding the Smith Jones example. The dictionary definition of entire should be enough to show that this is an error. I could also add that following your misinterpretation that "entire" meant any part of the paragraph, one could say that any single sentence alone is original research, or any phrase, etc., which of course would also be misinterpretations of the comments in WP:SYN regarding the example there. --Bob K31416 (talk) 15:06, 11 February 2009 (UTC)

Smith & Jones example

Ok, this is what we get when we include the whole package:

Argument 1 (simple deduction):

A = Jones claims that he consulted the original sources. [properly referenced]
B = The Harvard Writing with Sources manual requires citation of the source actually consulted. [source does not comment on dispute]

— Conclusion: (conclusion can be inferred if it is not stated)

C = If Jones's claim that he consulted the original sources is false, this would be contrary to the practice recommended in the Harvard Writing with Sources manual. [source does not comment on dispute]

and

Argument 2 (simple deduction):

A = Plagiarism is defined by the Harvard manual as using a source's information, ideas, words, or structure without citing them. [source does not comment on dispute]
B = The Harvard Writing with Sources manual requires citation of the source actually consulted. [source does not comment on dispute]

— Conclusion: (conclusion can be inferred if it is not stated)

C = The Harvard manual does not call violating this rule "plagiarism". [source does not comment on dispute]

and

Argument 3 (simple deduction):

A = If Jones's claim that he consulted the original sources is false, this would be contrary to the practice recommended in the Harvard Writing with Sources manual. [source does not comment on dispute]
B = The Harvard manual does not call violating this rule "plagiarism". [source does not comment on dispute]

— Conclusion: (conclusion can be inferred if it is not stated)

C = If Jones's claim that he consulted the original sources is false, this would not be contrary to the Harvard manual's definition of plagiarism. [source does not comment on dispute]

Or, instead of arguments 2 & 3 above:

Argument 2 (simple deduction):

A = Jones claims that he consulted the original sources. [properly referenced]
B = The Harvard Writing with Sources manual requires citation of the source actually consulted. [source does not comment on dispute]
C = The Harvard manual does not call violating this rule "plagiarism". [source does not comment on dispute]
D = Plagiarism is defined by the Harvard manual as using a source's information, ideas, words, or structure without citing them. [source does not comment on dispute]

— Conclusion: (conclusion can be inferred if it is not stated)

E = If Jones's claim that he consulted the original sources is false, this would not be contrary to the Harvard manual's definition of plagiarism. [source does not comment on dispute]

--Phenylalanine (talk) 12:21, 12 February 2009 (UTC)

Thank you, but first I would like a response to the point that I made. It's the point regarding the fact that you misinterpreted the statement in WP:SYN when you claimed that the quote from WP:SYN that you gave, said that the two sentences alone were original research. Do you agree now that what you quoted, viz. "This entire paragraph is original research..." was not referring to parts of that paragraph taken alone, like you tried to do by taking those two sentences alone? Do you recognize that now? --Bob K31416 (talk) 14:15, 12 February 2009 (UTC)

I should have said "This entire paragraph [including statements "B and C"] is original research...". The policy states: "This entire paragraph is original research, because it expresses the editor's opinion that, given the Harvard manual's definition of plagiarism, Jones did not commit it." The part I focused on in the essay does express the editors opinion, and the rest of the Smith & Jones example just goes off on an irrelevant tangent in my opinion (for example, if I copy part of someone's work on wikipedia, what difference does it make if I actually read the material copied or not...). "This entire paragraph is original research" means that all original deductive arguments (A,B, therefore C) within it are OR; and we can still consider these arguments individually as OR in accordance with the general WP:SYN principle. Cheers, Phenylalanine (talk) 02:01, 13 February 2009 (UTC)

Adding "including" to the brackets doesn't help. It is still a false implication that WP:SYN says that B and C without the rest of the paragraph is original research. Using your bracket approach of making a false implication, one could write, for example, "This entire paragraph [including statement "B"] is original research...". The false implication here is that WP:SYN says that statement "B" alone is original research.

Please note that you are definitely allowed to make up your own valid examples. But you are not allowed to make false implications about what WP:SYN says, for example by your bracket method. --Bob K31416 (talk) 04:15, 13 February 2009 (UTC)

Again, it's not a false implication. The policy says "the entire paragraph", not "only the entire paragraph", and the Smith & Jones example serves to illustrate the general WP:SYN principle (A, B, therefore C); so we should interpret the paragraph in accordance with that principle. Cheers, Phenylalanine (talk) 10:03, 13 February 2009 (UTC)

You seem to have a different understanding of the meaning of "entire" than the definition that I quoted from the dictionary. Also, I note that you didn't respond to the point that I made that using your reasoning, one comes up with the false implication that WP:SYN says that statement "B" is original research. Here's the point again:

Using your bracket approach of making a false implication, one could write, for example, "This entire paragraph [including statement "B"] is original research...". The false implication here is that WP:SYN says that statement "B" alone is original research.

Please respond to this point. Also, please note that I have put says in bold type to indicate that the issue is what WP:SYN says. Thank you.--Bob K31416 (talk) 10:48, 13 February 2009 (UTC)

Note that in the essay I mention "statements B and C" (then I conclude that "statement B" should also be considered original synthesis). All I'm saying is that there are multiple instances of synthesis (A, B, therefore C - as defined by WP:SYN) within the paragraph (several original logical arguments), and this does not contradict the fact that the "entire paragraph is original research) — which means that all components are original research within the context of the paragraph; this does not necessarily mean that, if you remove one or two components, automatically the paragraph no longer is original research. My interpretation of the Smith & Jones example is in line with the definition of synthesis, so there is no false implication. The Smith & Jones example should not be interpreted in isolation, but within the context of the WP:SYN section. Cheers, Phenylalanine (talk) 12:29, 13 February 2009 (UTC)

Unfortunately, you may not be understanding my point that I have tried to make numerous times in this discussion. It's not whether or not you are giving correct examples of synthesis, it's that you are attributing to WP:SYN something that it does not state, but rather something that is your opinion.

I hope you have found some of this long discussion useful. I'll be ending my participation now. Good luck. --Bob K31416 (talk) 13:59, 13 February 2009 (UTC)

I agree that I am attributing to WP:SYN something that is my opinion. Thank you for your helpful comments. Cheers, Phenylalanine (talk) 22:42, 13 February 2009 (UTC)

^
Definitions of primary sources:
- The University of Nevada, Reno Libraries define primary sources as providing "an inside view of a particular event". They offer as examples: original documents, such as autobiographies, diaries, e-mail, interviews, letters, minutes, news film footage, official records, photographs, raw research data, and speeches; creative works, such as art, drama, films, music, novels, poetry; and relics or artifacts, such as buildings, clothing, DNA, furniture, jewelry, pottery.
- The University of California, Berkeley library offers this definition: "Primary sources enable the researcher to get as close as possible to what actually happened during an historical event or time period. Primary sources were either created during the time period being studied, or were created at a later date by a participant in the events being studied (as in the case of memoirs) and they reflect the individual viewpoint of a participant or observer."

[1] Definitions of primary sources:
The University of Nevada, Reno Libraries define primary sources as providing "an inside view of a particular event". They offer as examples: original documents, such as autobiographies, diaries, e-mail, interviews, letters, minutes, news film footage, official records, photographs, raw research data, and speeches; creative works, such as art, drama, films, music, novels, poetry; and relics or artifacts, such as buildings, clothing, DNA, furniture, jewelry, pottery.

The University of California, Berkeley library offers this definition: "Primary sources enable the researcher to get as close as possible to what actually happened during an historical event or time period. Primary sources were either created during the time period being studied, or were created at a later date by a participant in the events being studied (as in the case of memoirs) and they reflect the individual viewpoint of a participant or observer."

[2] The University of Nevada, Reno Libraries define primary sources as providing "an inside view of a particular event". They offer as examples: original documents, such as autobiographies, diaries, e-mail, interviews, letters, minutes, news film footage, official records, photographs, raw research data, and speeches; creative works, such as art, drama, films, music, novels, poetry; and relics or artifacts, such as buildings, clothing, DNA, furniture, jewelry, pottery.

[3] The University of California, Berkeley library offers this definition: "Primary sources enable the researcher to get as close as possible to what actually happened during an historical event or time period. Primary sources were either created during the time period being studied, or were created at a later date by a participant in the events being studied (as in the case of memoirs) and they reflect the individual viewpoint of a participant or observer."

[1]