Jump to content

Wikipedia talk:No original research/Primary v. secondary sources discussion

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


The History of the Conflict

Is the addition of the distinction between primary and secondary sources a new addition? I don't remember it being there before. By the way, it's utterly awful. the change creates confusion and opens the door for crack pots who can argue semantics. Dire. --I'll bring the food 00:35, 27 August 2006 (UTC)

no, it is not a new distinction. The very first version (December, 2003, byt User:Tarquin and based on discussion on the list-serve) stated that original research such as new scientific theories is not allowed — in other words, the policy in its original formulation prohibited not just the creation of new fact but theories about facts. In February 2004 User:Reddi added a specific clarification that the issue was new theories not new facts (everyone I believe took it for granted that introducing new and undocumented facts was just unimaginable; the focus was on contributors interpreting accepted facts). Later that month Reddi added the comment that Wikipedia should not be a primary source, and then a few edits later made the distinction between primary and secondary sources — in other words, the distinction was implicit in the first version and within two months was made explicit. At that point, "primary source" was undefined, but the definition of secondary source as it appears today was established. There was however ongoing debate as to whether Wikipedia should be a secondary or tertiary source. Editors agreed that the cornerstone of the policy is:
The phrase "original research" in this context refers to untested theories; data, statements, concepts and ideas that have not been published in a reputable publication; or any new interpretation, analysis, or synthesis of published data, statements, concepts or ideas that, in the words of Wikipedia's founder Jimbo Wales, would amount to a "novel narrative or historical interpretation".

The current definition of primary sources and the basic framework for when primary sources can and cannot be used was hammered out in March 2005, and has been stable since that time. Slrubenstein | Talk 15:19, 27 August 2006 (UTC)

ROOC 1 (27 Aug 2006). By way of a very brief response to KB's query, I think that it's fair to say that the crux of the dispute is about the language that SLR recently added to the main page, most especially this sentence:

Research that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources.

JA: It is my impression that many editors regard this as excessive originality in policy-making on SLR's part. Jon Awbrey 19:22, 27 August 2006 (UTC)
My conflict was with the precursor version which stated "draws" instead of "predominantly draws". To me, the use of "predominant" allows a case-by-case analysis and resolution process, where "draws" appears to be too restrictive, as it can be read to exclude primary sources entirely. Hopefully, my characterization is accurate, when I say that the articles I myself create, are a mixture of both primary and secondary sources. For example, a secondary source might state "vanished from history after 1950" and yet a newspaper article from 1955 might state "so-and-so threw a gala affair at which the Governor of New York and his wife attended". There is room for both quotes in the article. The newspaper eye-witness report (presumably eye-witness) would be a primary source in this case. Wjhonson 02:37, 28 August 2006 (UTC)
I think Jayjg hit the nail on the head with : "Editing using primary sources is one of the most typical ways people insert original research into articles. It's fine to quote a primary source when one is being completely non-analytical, but as soon as one starts evaluating the source material one must avoid using primary sources to back up that evaluation. Slrubenstein's edit made that clearer. Jayjg (talk) 20:46, 24 August 2006 (UTC)" Is there something we can add to "Research that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources" so that Jayjg's point is clear and the sentence is less likely to be used as a weapon against proper uses of primary sources? Or is that impossible for our esteemed wordsmiths? WAS 4.250 04:15, 28 August 2006 (UTC)

For clarity, I am reproducing the section in question.

  • Primary sources present information or data, such as archeological artifacts; film, video or photographs (but see below); historical documents such as a diary, census, transcript of a public hearing, trial, or interview; tabulated results of surveys or questionnaires, records of laboratory assays or observations; records of field observations; literary works such as poems, scripts and screemplays, or novels.
  • Secondary sources present a generalization, analysis, synthesis, interpretation, or evaluation of information or data from other sources.
Original research that creates primary sources is not allowed. Research that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources. However, where an article (1) makes only uncontentious descriptive claims the accuracy of which is easily verifiable by any reasonable adult without specialist knowledge, and (2) makes no analytic, synthetic, interpretive, or evaluative claims, a Wikipedia article may be based entirely or primarily on primary sources (examples would include apple pie or current events). These are relatively rare exceptions and contributors drawing predominately or solely from primary sources should be exceptionally careful to comply with both conditions.
Research that consists of collecting and organizing information from primary and secondary sources is, of course, strongly encouraged. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia.
It is essential to the quality and integrity of Wikipedia that all articles be based on information collected from verifiable sources. Wikipedia articles include material on the basis of verifiability, not truth. That is, we report what other reliable sources have published, whether or not we regard the material as accurate. In order to avoid doing original research, and in order to help improve the quality of Wikipedia articles, it is essential that any primary-source material, as well as any generalization, analysis, synthesis, interpretation, or evaluation of information or data, has been published by a reputable third-party publication (that is, not self-published) that is available to readers either from a website (other than Wikipedia) or through a public library. It is very important to cite sources appropriately, so that readers can find your source and can satisfy themselves that Wikipedia has used the source correctly.
In some cases, there may be controversy or debate over what constitutes a legitimate or reputable authority or source. Where no agreement can be reached about this, the article should provide an account of the controversy and of the different authorities or sources. Such an account also helps ensure the article's neutral point of view.

A number of people have expressed their support for or acceptance of this version as it stants: Slrubenstein, SlimVirgin, Wjhonson, Jossi, Felonious Monk and I think Jayjg. If there continue to be specific objections, let´s try to sort them out one at a time.

Threshold for NOR in paraphrases

I have one remaining issue with the current NOR policy. When paraphrasing or summarizing a source, the skills used might be those of any good reader and writer. But in technical material, somewhat more specialized skills might be used. Units of measure might be converted. Words like best or worst might be inserted where those words are only implied in the source, in combination with measurements that show which is best or worst. Numbers might be read from graphs, or graphs might be made from lists of numbers. All of these changes are arguably not original research and should be allowed. The process could be carried further, for example, by extrapolating the data from the source to new situations; that would be original research.

For most sources, the threshold is not explicitly set, and it would be up to a consensus of the editors for an article to decide what was or was not OR. But for articles based primarily on primary sources, the threshold is now set: any reasonable adult without specialist skills must be able to verify that the article is an accurate representation of the source. This would prevent some paraphrases that really involve no original research, because a person who lacks the skills to make a graph or convert units cannot verify the accuracy of the paraphrase. --Gerry Ashton 16:19, 27 August 2006 (UTC)

Gerry, I think the issue here is that when different people were working on the policy, they were going by two kinds of "primary source" - some are themselves verifiable (i.e. census data widely available on the web) but others are not verifiable as sources, e.g. a laboratory assay which is unpublished and uncirculating. I think people put in the word "accuracy" to cover those cases where the primary source (meaning, a written document) itself is not verifiable (meaning, available to the pbulic) but the accuracy of whatever claim is being made (I dunno, apple pies have apples in them?) is in fact easily comfirmable. I am not sure what the most elegant solution to the problem is, but I hope I have shed some light on the nature of the problem. Do you see what I mean? The word "accuracy" was originally introduced I think as a loophole, to make the policy more liveral and not more restrictive - what happens if the "primary source" is not readily available to the public (like, my grandma´s recipe for apple pie)? I think the confusion comes from using the word "verifiable" in two different ways, one meaning "a source readily available to th public" and the other meaning "a fact obvious to anyone." Maybe it would suffice to change "verifiable" in this clause to "confirmable?" Slrubenstein | Talk 16:32, 27 August 2006 (UTC)
The issue of common knowledge (about apple pies for instance) that can't be readily found in a written source can be dealt with, in part, by considering commonly available artifacts to be primary sources. If for some reason I needed to write that the prongs of 110 volt power plugs in North America are 13 mm apart, I could go searching through the standards, and perhaps find that I'd have to buy a $50 standard from the IEEE, which might or might not contain the data. Or, I could just grap a ruler and a power plug and measure. Since power plugs are readily available to almost everyone in Wikipedia reader in North America, and so are rulers, the power plug itself can serve as a primary source. --Gerry Ashton 16:49, 27 August 2006 (UTC)

I don't really see much of a conflict here. On the one hand, the policy refers to the ability of a "reasonable adult without specialist skills" to verify the information. On the other hand, Gerry gives the example of changing units, paraphrasing, summarizing, etcetera (in ways that are clearly not a novel synthesis), which require some ability to read and understand the source material. However, I wouldn't consider the ability to convert units, read a graph, etcetera, to be a "specialist skill". Those are skills that any reasonable adult would have to possess in order to even begin to evaluate an such an article.

It is inevitable that, the more technical the subject of an article is, the more restricted the set of readers able to evaluate it becomes. This does not mean that articles should be written only for "specialists" — a specialist is someone who is an expert in a particular field, someone qualified to write academic papers in that subject, and I fully agree with the policy that articles should be verifiable by non-specialists. But, for example, if an editor is to verify an advanced physics or mathematics article, they should have some general knowledge physics or mathematics in order to make any sense of it. A "reasonable adult" will not try to evaluate an article on character theory if they haven't taken a math course since high-school algebra 20 years ago.

Ultimately, a policy like this has to be written in broad language that requires some common-sense interpretation. It can't be a precise legal document. I don't see any big problem with the current language, but if it would help to clarify you could change "any reasonable adult" to "any reasonably informed adult".

—Steven G. Johnson 02:13, 28 August 2006 (UTC)

Exactly who is this reasonable adult? It has been suggested that .pdf files and excel files should not be allowed in Wikipedia since the the "average" reader may not know how to download a free program to open them, for example Adobe Acrobat Reader. This would exclude most peer-reviewed articles that are almost always available on the net as .pdf files.Ultramarine Ultramarine 02:41, 28 August 2006 (UTC)
Please don't change the subject. There are good reason for excluding closed formats like .xls from being used on Wikipedia that have nothing to do with requiring specialist knowledge. And there is a strong consensus that linking to .pdf journal articles is perfectly fine; this is a straw man. —Steven G. Johnson 03:33, 28 August 2006 (UTC)
What are those reasons? This would require deleting many of the charts and graphs in Wikipedia. Note that also pdf is a closed format. .xls files can easily be opened in Open Office.Ultramarine 04:16, 28 August 2006 (UTC)
Sorry, I'm not going to get into an off-topic debate with you; that would be a waste of time. —Steven G. Johnson 04:30, 28 August 2006 (UTC)
I hope people would interpret "specialist" as Mr. Johnson does. However, the word specialist has been diluted in U.S. industry and the U.S. armed forces to describe those with a fairly low level of technical training. People who have been exposed to these diluted meanings might very well think reading a graph (especially if the scale is logarithmic) is a specialist skill. I've tried to think of a formulation that would expect an evaluator to generally understand the subject, without requiring that the evaluator have an advanced degree in the subject, but I can't think of anything.
I would also like to see a formulation that does not include the phrase "reasonable adult", because if I ever wanted to say that a paraphrase should remain in an article over the objections of editor X, I would not want to refer to X as an unreasonable adult. --Gerry Ashton 02:43, 28 August 2006 (UTC)
I don't think that you're ever going to find a wording that prevents unreasonable people from complaining. In cases of conflict, the solution on Wikipedia is always going to be the same: you have to convince other editors (or at least a majority) that you aren't adding any new interpretations or drawing novel conclusions, but are simply describing what is already in a reputable source. —Steven G. Johnson 03:33, 28 August 2006 (UTC)

So far, the hypothetical situations like unit conversions don't bear much relation to reality for me—Wikipedians make unit conversions all the time with zero controversy. It would be clearer what you want if you could propose a specific wording that you think is better. If you can't propose an improvement, then this discussion is pointless. —Steven G. Johnson 03:14, 28 August 2006 (UTC)

I made a specific suggstion above, but no one responded. I thought Gerry would at least respond to my suggestion, one way or the other, since I was responding to hiw criticism. Is Gerry commited to a constructiv discussion or are we just waisting our time here, as Stevenj observes? Slrubenstein | Talk 16:49, 28 August 2006 (UTC)
If you are refering to changing verifiable to confirmable, I don't think that would help. As you describe the history of the page, this section intertwines confirming whether quotes, paraphrases, and summaries accurately represeent sources with including sourceless information if that information is easily confirmable. Since the heading is "Primary and secondary sources", if the people who composed this section thought they were writing about unsourced information, they were wrong. The heading disallows any interpretation of this section refering to sourceless information. --Gerry Ashton 18:15, 28 August 2006 (UTC)
I was referring to that. Okay, so you think it is inadequate. What do you propose instead? Please bear in mind that this policy was not written by a collective of people acting in unison at one time. Like other things at Wikipedia many different people contributed over time. So we shouldn´t be surprised when every once in a while a sharp reader catches an unncessary or confusing ambiguity or contradiction. Instead of getting churlish, why don´t we look for solutions. I proposed one with good intentions. Can´t you do likewise?Slrubenstein | Talk 18:47, 28 August 2006 (UTC)

Secondary ≠ Better

JA: The skinny on my objection to Rubenstein's inventions is that they imply that secondary sources are inherently better or higher quality in some insinuated but unidentified way. Nobody who reads any philosophy at all would ever dream of saying that. No philosopher (Phil1) has ever written about the work of another philosopher (Phil0) in a way that suffices to replace Phil0's work, even though it is not unheard for Phil1's ilk to suggest that. Ergo, there can be no sensible policy from which the reader might infer that secondary sources are somehow preferred. Jon Awbrey 12:32, 28 August 2006 (UTC)

I agree; worse, it's off-topic and may occasionally conflict with policy - see my recent discussion on the Talk page. Harald88 19:27, 21 September 2006 (UTC)
Saying that secondary sources are preferred sources for an encyclopedia does not imply that they are better in an absolute sense. The policy should say what is good for Wikipedia, no more, no less. It seems like most editors here agree with this (including me), so it's inappropriate for you to personalize the issue as "Rubenstein's inventions".
(By the way, this is hardly novel to Wikipedia. I've written for technical print encyclopedias, and they also recommended citations to textbooks, reviews, etc. rather than primary sources where possible. The case for Wikipedia is even stronger because of its openness.) —Steven G. Johnson 15:26, 28 August 2006 (UTC)

JA: Here's a recipe for writing a low-quality article:

  1. Read a single secondary, tertiary, or popularization-type work on a subject.
  2. Write the article.
  3. Do not check its statements against other 2ndary and 3rdary sources.
  4. Do not check its statements against primary sources.

JA: Maybe this practice is pandemic in WP, but it does not conform to:

  1. Standards of good reporting.
  2. Standards of good scholarship.
  3. Standards espoused by the WP:Policies of NOR, NPOV, and VER.

JA: It's really no more complicated than that. Jon Awbrey 16:18, 28 August 2006 (UTC)

Here's another recipe for writing a low-quality article: read only primary sources and, because you aren't a specialist, misinterpret them or fail to take them into the proper context. There are plenty of ways to write a terrible article, but an editor without specialist knowledge is generally going to be better off reading and summarizing secondary sources most of the time. Relying on secondary sources is a standard of good scholarship/reporting when working outside your field of expertise. Even in the rare case where the editors are specialists (in which case your scenario does not apply), references to secondary sources such as textbooks or reviews are much more likely to be useful to readers. Furthermore, an emphasis on secondary sources helps to steer even expert editors away from the temptation of original research. —Steven G. Johnson 18:32, 28 August 2006 (UTC)
As I explained on the Talk page, editors should as much as possible use the original sources - which may be primary as well as secondary. It's moslty a subject for WP:V. Harald88 19:27, 21 September 2006 (UTC)
Gosh, Steve, it sounds like you have an awful lot of experience working here at Wikipedia and have a good idea about how to assess realistic, constructive policies! Thanks! Slrubenstein | Talk 18:58, 28 August 2006 (UTC)
Why can't we say both in the policy? As near as I can tell everyone agrees that secondary sources are useful when used right and primary sources are useful when used right. Both can be misused. I think a single added sentence indicating somehow that a primary sources is not necessarily worse than a secondary source would be useful when someone is properly using a primary source and someone else wishes to use NPOV as a weapon in replacing it with a secondary source without investigating if it is or is not a correct use of a primary source. And if people say it already says that then what could hurt with a single sentence making it more clear that that is the case. WAS 4.250 23:21, 28 August 2006 (UTC)
Because secondary sources are preferred, for the reasons I explained (and as others have explained...over and over). The policy does not say that primary sources are necessarily "bad". Just that secondary sources are often better for an encyclopedia like Wikipedia. Please go back and re-read what I wrote above. —Steven G. Johnson 00:01, 29 August 2006 (UTC)
It sounds like you are saying secondary sources are generally but not always preferred or else the "often" in "secondary sources are often better for an encyclopedia like Wikipedia" should be "always". This is precisely the point that nees to be more clear. WAS 4.250 01:09, 29 August 2006 (UTC)
I'm always suspicious of these "hypothetical" cases. If someone is misquoting a secondary source, or using as a "weapon" a secondary source that is obviously contradicted by other reputable sources, it should be no problem to convince other editors of that. If you can't, it might be a sign that your use of the primary source is not as "proper" as you think. No wording of the policy is going to turn unreasonable people into reasonable ones. —Steven G. Johnson 00:07, 29 August 2006 (UTC)
I'm always suspicious of these "hypothetical" cases too. We are talking about real cases. Use of primary sources by me concerning H5N1 which reflect proper use of primary sources (no one has said "boo" to me about it, which is good) and Charles Peirce related articles edits by Jon which he seems to be indicating he has had some difficulty with (I find his edits on those page quite good altho I haven't specifically evaluated the primary source angle in detail, I am broadly familiar with the nature of the sources available). WAS 4.250 01:09, 29 August 2006 (UTC)
As for "it should be no problem to convince other editors of that" can you honestly say you believe Jon has been effective at the skill set people use for the task of convincing other editors? Maybe its not so trivial for everyone. Just maybe each of us has differing levels of abilities. Maybe? WAS 4.250 01:14, 29 August 2006 (UTC)
Let me understand you. At H5N1 you have used primary sources in a way you feel is proper, and no other editor has challenged you. This means that the policy is a problem, why? As for the Peirce article, JA was challenged for allegedly including original research material amounting to a novel synthesis and analysis of Peirce's work taken from JA's own PhD thesis, which (apparently) was not published in reputable peer-reviewed journals. On its face, this objection seems entirely reasonable to me. Whether the synthesis is correct is irrelevant to Wikipedia; the standards for an encyclopedia are not the same as the standards for a thesis or a journal article. —Steven G. Johnson 15:35, 29 August 2006 (UTC)

(<---) I agree with everything you said and everything I understand you to have implied. The H5N1 example is useful to illustrate that sometimes primary sources are useful and the policy should not say otherwise, as it right now indeed does not, even if it could be more clear in that regard. I don't have a problem with the current wording, I merely have a problem with seeing why it is a problem to add a few words to make clear what what all agree is the case. Perhaps if Jon had better people skills, he might have been more successful in obtaining his objectives with regard to Peirce which as you point out had clear problems on their face. But the bottom line for me here is why not offer to add a couple of words that clarifies at best and is redundant at worse rather than go through all this mess? WAS 4.250 22:38, 29 August 2006 (UTC)

I have no problem with clarifying the policy, if a better wording can be suggested. (Perhaps I misunderstood your remarks; I thought you wanted to change the policy so that both secondary and primary sources were treated equally.) Do you have a specific wording in mind? —Steven G. Johnson 00:49, 30 August 2006 (UTC)
Neither of us has a problem with clarifying the policy. Neither of us wishes to change the policy. I have no specific wording in mind. I only suggest that seeking such a specific wording does not qualify as trolling and should be dealt with as a legitimate concern rather than as an act to be despised. But Jon needed to be spanked; he was not paying attention to the results of his behavior (pissing people off). I hope this can all be worked out ok. WAS 4.250 04:00, 30 August 2006 (UTC)

We prefer the use of 2ndry sources over primary ones for extra-editorial breadth, balance, expertise, and to ensure against no original theories. The current formulation can probably be stated more simply (and without all the Howevers, those are painful to read!). I'll give it some thought. El_C 06:35, 31 August 2006 (UTC)

I conclude that the claim that "secondary sources are preferred" is not consensual - and not surprisingly so. Even in the discussion below just a hand-full of editors discussed that option among other alternatives that don't conflict with other policies. Harald88 19:27, 21 September 2006 (UTC)

H5N1 Article relies on Secondary, not Primary Sources

I looked at this example, and I don't follow why this is one.

In the scientific context, primary sources would be the raw experimental data, preprints, conference discussions, etc. On quick inspection, the article relies on peer-reviewed articles, science press reports, and so on which are secondary sources. Whether the government-sponsored papers are primary or secondary is a semantic debate -- they are proper for inclusion either way. There may be an occasional reference to other primary sources, but it doesn't jump out at me -- there is no objection current or proposed to using primary sources as leaven in an article relying on secondary sources.

What am I missing here? Robert A.West (Talk) 16:33, 30 September 2006 (UTC)

I dispute that the peer-review process makes a secondary source out of a primary one. A primary source is one which presents new data for the first time. As such, the published paper, albeit peer-reviewed is still a primary source. Wjhonson 17:29, 30 September 2006 (UTC)
The H5N1 and Flu series of articles relies on both primary and secondary sources and is an example of correct use of primary sources (as you discovered) and not an example of incorrect use of primary sources. The point of the example is to make sure primary sources are not thrown out or left out of the list of acceptable (even prefered on occation) sources. For example, the H5N1 situation is literally a new thing on the Earth with no known precedent for its current behavior so that much information about flu is out of date and some secondary sources rely on experts who haven't kept up. Primary sources were occasionally necessary to clarify a confused, changing, and very important topic. Researchers are continuing to claim "new" results in the media for facts published in primary sources literally years ago (example: immune system overreaction in spanish flu and H5N1). WAS 4.250 17:45, 30 September 2006 (UTC)
Note that guideline Wikipedia:Reliable sources says: "The popular press generally does not cover science well. Articles in newspapers and popular magazines generally lack the context to judge experimental results. They tend to overemphasize the certainty of any result, for instance presenting a new experimental medicine as the "discovery of the cure" of a disease. Also, newspapers and magazines frequently publish articles about scientific results before those results have been peer-reviewed or reproduced by other experimenters. They also tend not to report adequately on the methodology of scientific work, or the degree of experimental error. Thus, popular newspaper and magazine sources are generally not reliable sources for science and medicine articles. What can a popular-press article on scientific research provide? Often, the most useful thing is the name of the head researcher involved in a project, and the name of his or her institution. For instance, a newspaper article quoting Joe Smith of the Woods Hole Oceanographic Institution regarding whales' response to sonar gives you a strong suggestion of where to go to find more: look up his work on the subject. Rather than citing the newspaper article, cite his published papers." WAS 4.250 18:02, 30 September 2006 (UTC)
At least we have a sharp, well-defined point of disagreement. Lab notebooks are clearly a primary source, as are preliminary notes put together on the subject. A report in Science Digest is clearly a secondary source. What about a peer-reviewed article by the researcher himself? I would say that it is closer to a secondary source than a primary, both in nature, and in its utility: it provides expert interpretation of the raw facts, and has gone through a fact-checking process that is not under the control of the author.
You state that a source where an idea appears for the first time is primary. That is not the case. The letters between Lise Meitner and Otto Hahn are primary sources. The first paper to suggest that she was unfairly denied credit for the discovery of fission is a secondary source. Robert A.West (Talk) 20:46, 30 September 2006 (UTC)
Not in the context of this wiki. A primary source is merely a source which presents, for the first time, an idea. One source can be primary for statements A, B and C and secondary for statements D, E, and F. If a paper says "She was unfairly denied credit... and she told me so." Then it's secondary for the opinion, but primary for the reporting of what she said.Wjhonson 22:12, 30 September 2006 (UTC)
The Meitner-Hahn example skips over the publications that this discussion is focusing on. Sure letters are primary sources. The first claim that Meitner was unfairly denied credit for fission is apt to be secondary, since a good author would have reviewed the letters and papers published by Meitner-Hahn, as well as related papers from other authors. The question for this discussion would be whether the papers published by Meitner and Hahn which report their results for the first time are primary sources or secondary. My take on it is that a secondary source is one based on previously published primary sources. A published paper based on the primary sources (such as lab notebooks) that are privately held among the authors and close associates is still a primary source. --Gerry Ashton 22:51, 30 September 2006 (UTC)
That is not the distinction on the project page. To quote, "Secondary sources present a generalization, analysis, synthesis, interpretation, explanation or evaluation of information or data from other sources." I read that unequivocally to mean that a peer-reviewed paper is a secondary source derived from the lab results, which are the primary source. Robert A.West (Talk) 22:53, 30 September 2006 (UTC)

Critical Points

JA: GTB, up until 15 Aug 2006 the WP:NOR policy had remained pretty much the same for as long as I personally had been watching it, since 20 Dec 2005 at least.

  1. On or about 15 Aug 2006 or so, Kim Bruning tried to excise or weaken the clauses about "non-negotiability" in light of his/her views about WP:NBD. These were fairly minor skirmishes compared to what happened next.
  2. The shot that ignited the civil war in progress seems to have been this edit by Slrubenstein:

JA: Slrubenstein's edit, along with all of its subsequent variations, would constitute a major alteration of long-standing policy on Original Research. I think that it's fair to say that there can be no consensus for making this change. The supporters of this policy modification have, however, simply refused to hear the objections to this change, and some have even tried to manipulate the normal procedures of talk page use in order to set aside the continuing protest.

JA: That is how I see the current situation. Jon Awbrey 02:12, 3 September 2006 (UTC)

Thank you for that clear summary, Jon. I'm going to look a bit more deeply into this. I do feel the paragraph being suggested is an improvement over the current version, and it appears that the issues you're raising go far beyond the fairly minor edit that's being suggested. Would you agree with that? -GTBacchus(talk) 02:17, 3 September 2006 (UTC)

JA: I can't be sure, as we need a better way of indexing the various versions in play. By "current version" I would assume that you mean the version that is currently locked down on the main page. This is one of the subsequent variations on an original theme of Rubenstein, and it still amounts to a radical re-invention of previous policy. For "paragraph being suggested" I do not know if you mean a variant on this page or elsewhere. FeloniousMonk unilaterally changed the venue for discussing the proposed alterations in the status of primary vs. secondary sources to here:

JA: I started a longitudinal study of the changes in the policy page since its inception here:

JA: Will try to take a day off now. Good weekend, Jon Awbrey 03:04, 3 September 2006 (UTC)

JA: Okay, I just now read the assertions being made in the section above, and that forces me to say one more thing. It is disingenuous of anyone to pretend that no specific objections to the locked down version or its cosmetic variants have yet been put on the table. The record, no matter where somebody chooses to relocate it, will continue to show that the same objections have been made since the time that the changes in question were introduced. The essence of the specific recommendations have always been: "Put it back the way it was before 15 Aug 2006." Let us hope that this is specific enough for all to understand. And now I'm really outa here. Jon Awbrey 03:22, 3 September 2006 (UTC)

It appears to me that the change being proposed above is an edit to the "Citing oneself" section, and doesn't really touch the "Primary and Secondary sources" issue that you're concerned with. It may be advisable to get on the same page regarding both of those sections before unprotecting. Have a good weekend, anyway. -GTBacchus(talk) 03:33, 3 September 2006 (UTC)
Precisely. One way to resolve disputes is to handle them one at a time, or separately. A separate talk page was created to discuss primary versus secondary issues, and people have been and can debate that issue, and work towards an agreement, there. Here however we have been focusing on the issue of expert authors. Let´s do one thing at a time - that is the way to proceed further. Only trolls who wish to see no progress at all keep mixing up different issues in order to prevent or disrupt any progress at all.Slrubenstein | Talk 14:18, 3 September 2006 (UTC)

JA: Yes, "staying on one page" (SOOP) is rather a problem here, and I don't see that creating a complex of subpages has really helped the SOOP situation all that much, nor the "one can of worms at a time" (1COWAAT) situation neither. Nor does it serve much constructive community purpose to be constantly attributing trollitude, Dude, to anyone who now and then gets bewildered by the situation. Jon Awbrey 15:54, 3 September 2006 (UTC)

Yeah, I really don't see how characterizing anybody as a troll is actually going to lead to a solution more smoothly or quickly. Anyway, Morven gave me the green light to unprotect. Let's go ahead and implement the change to the "Citing oneself" section, and then talk about how to handle the Primary v. Secondary sources issue. I suggest that none of us dive in and start making Primary v. Seconday edits until we've achieved a bit of understanding through discussion. In particular, if anybody starts reverting to previous versions, I'll request reprotection immediately. Let's make that "Citing onself" edit, and make sure it's stable, and then we'll address the other can o' worms. Sound good? -GTBacchus(talk) 18:54, 3 September 2006 (UTC)
Ok, the protection is lifted, and the edit in place - let's see what happens. Meanwhile, I hear there's some disagreement as to Primary v. Secondary sources... Jon Awbrey seems to be claiming that Slrubenstein has made a substantial and detrimental change to the policy as regards sources. The edit in question... is actually scores of different edits by now, but this diff appears to capture the issue pretty well. Slrubenstein, can you say something about how you see this issue? Or, should I be asking this question on the subpage? -GTBacchus(talk) 19:20, 3 September 2006 (UTC)

Let´s move forward one step at a time

step 1: What counts as a primary source?

Here is one change I made which was reverted prior to protection. I would like to put it back in. I would like to know if there are any objections. To the current list of primary sources, I would like to add this: "literary works such as poems, scripts and screemplays, or novels." Any objections? Slrubenstein 20:24, September 8, 2006 (UTC)

I support this addition, and the slow approach. -GTBacchus(talk) 20:43, 8 September 2006 (UTC)
I support this addition in principle, but I dislike the way the paragraph intertwines the nature of a work with the medium. I would change "film, video, photographs (but see below)" to just "photographs (but see below)" and would would alter the proposed change to say "literary works such as poems, scripts and screenplays, novels, or fictional motion pictures, videos, and television programs."
My version would not explicitly cover raw moving images, such as might be used to record a scientific experiment, security camera videos, and the like; I don't know if we need a clause to cover that or not. --Gerry Ashton 20:53, 8 September 2006 (UTC)

Gerry, I like your version except I would just drop the word "fictional." How does that sound? Slrubenstein | Talk 21:17, 8 September 2006 (UTC)

My purpose in putting the word fictional before "motion pictures, videos, and television programs" is to recognize that some programs in those media are secondary sources rather than primary sources. For example, the episodes of the Nova television series would qualify as secondary sources. --Gerry Ashton 21:30, 8 September 2006 (UTC)

Good point - I didn´t think of tht. Okay, we keep fictional. But I do believe we still need somehow to include the other kinds of raw video you mention. So, how about this:

Primary sources present information or data, such as archeological artifacts; photographs (but see below); historical documents such as a diary, census, video or transcript of surveillance, a public hearing, trial, or interviews; tabulated results of surveys or questionnaires; written or recorded records of laboratory assays or observations; written or recorded records of field observations; artistic and literary works such as poems, scripts and screenplays, novels, or fictional motion pictures, videos, and television programs.

I put your changes and mine in italics. Slrubenstein | Talk 21:42, 8 September 2006 (UTC)

I would drop the word script because it is covered well enough by screenplay; I believe screenplays are understood to be fiction while scripts apply equally to fictional and documentary plays and moving images. --Gerry Ashton 21:58, 8 September 2006 (UTC)

I disagree that screenplay includes script - script can apply to theatrical productions but screenplay does not. Also, I would just remove the word "fictional" since the whole clause is now "artictic or literary" which I certainly think subsumes "fictional," right? Slrubenstein | Talk 22:03, 8 September 2006 (UTC)

You are certainly right about screenplay not including plays, so script should stay after all. You are also right that "artistic" excludes moving pictures that are secondary sources, but a secondary source that reviews or critisizes another work could be considered literary, so I would write fictional rather than literary. I am also woried that the passage has so many and's, or's, commas, and semicolons that it could easily be misread. Perhaps we could make the passage look like this:

Primary sources

Primary sources present information or data, such as

  • archeological artifacts
  • photographs (but see below)
  • historical documents such as a diary, census, video or transcript of surveillance, a public hearing, trial, or interviews
  • tabulated results of surveys or questionnaires
  • written or recorded records of laboratory assays or observations
  • written or recorded records of field observations
  • artistic and fictional works such as poems, scripts, screenplays, novels, motion pictures, videos, and television programs.

Secondary sources

Secondary sources present a generalization, analysis, synthesis, interpretation, or evaluation of information or data from other sources.

My personal preference is for the older (i.e. current) format (semicolons). That said, i have no objection to Gerry Ashton´s proposal. Let´s see if one or two more people won´t weigh in. If someone else objects to the bullet-point format, Gerry, is it acceptable to you to stick to the current format? If others support Gerry´s approach, as I said, I won´t object. Let´s just get a little more feedback from others, and then one of you can make the change. Slrubenstein | Talk 22:40, 8 September 2006 (UTC)

If we do not adopt the bullet format, we must give exceptional care to all the and's, or's, commas, and semicolons. I wouldn't be able to say yea or nay to a new proposal in the existing format without having it completely spelled out. I also would like others to weigh in. --Gerry Ashton 22:49, 8 September 2006 (UTC)

Are we going to stress that any primary sources we use must have been published by a reliable source? SlimVirgin (talk) 02:46, 9 September 2006 (UTC)

I think that should be the next topic. Given the acrimonious and unproductive arguments that led to the page being protected, I think it makes sense to take each edit one at a time. right now we are discussing the definition of primary sources. I think we have a consensus, but let us be sure we do and then make the change. Then let us move on to the next topic. Slim, I think what you suggest is fine but I think it should be stated in a different paragraph, so can we disacuss this next *sorry my keyboard lacks a question mark( Slrubenstein | Talk 03:30, 9 September 2006 (UTC)

I'm enjoying some of your posts with the dodgy keyboard. ;-) SlimVirgin (talk) 05:10, 10 September 2006 (UTC)

I think the bulleted version is fine. SV does have a point; we want to be sure that nobody thinks they can use the arrowhead they found in their backyard, or, as happens occassionally, the manuscript they found in their attic, as a source. -- Donald Albury 12:23, 9 September 2006 (UTC)

For me that's the main issue, not whether something is a primary or secondary source. The main thing is that all our source material must have been published by a reliable source. After making that clear, we can define primary and secondary source, and say that preference is normally given to secondary sources, simply because primary sources can be hard to use properly, but that both are acceptable. I won't argue the point because there's been enough back and forth, but I just wanted to throw it in for consideration. SlimVirgin (talk) 05:10, 10 September 2006 (UTC)

I have no desire to weigh in or debate the issues that I'm now going to bring up. It's just that I see these issues raising their ugly head in various places and sooner or later they'll have to be dealt with, so if stuff in this policy is to be changed/rewritten, if should be with full understanding of consequences and possible interpretations.

  1. Reliable accessable source = the real world? (sky is dark at night, water is wet when warm, covered bridge is functional and not just for storage)
  2. DVD is a publication?
  3. broadcast is a publication?
  4. a rare book is not a publication? WAS 4.250 06:57, 10 September 2006 (UTC)
WAS's third point affects Sirubenstein's proposed definition of a primary source, in that the definition does not differentiate between television programs and motion pictures for which a transcript or recording is (or has ever been) available, and those for which a trancript or recording has never been available. --Gerry Ashton 17:38, 10 September 2006 (UTC)

I made an edit to the policy, puttingin Gerry´s list in buller-points, and I added concerning movies and such "whether in analogue or digital format." If this does not satisfy WAS´s concern about DVDs, I ask Gerry or WAS to make the appropriate correction. Personally, I think it is a non-issue. It is the contents not the format that is at issue here. Television programs and motion pictures counts as original research period, what difference does it make whether it is on 16 mm film, video, or dvd? However, if Gerry Ashton can find a way to make the phrasing more parsimonious, precise, clear, go ahead. Slrubenstein | Talk 19:21, 11 September 2006 (UTC)

There are three more edits I propose. I have NOT made thes edits put am laying them out, in order, here on the talk page first. Slrubenstein | Talk 19:22, 11 September 2006 (UTC)

Step 2: clarify our position concerning primary sources compared to secondary sources

Currently, the first two paragraphs following the descriptions of primary and secondary sources is:

Original research that creates primary sources is not allowed. However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged. All articles on Wikipedia should be based on information collected from published primary and secondary sources. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia.
In some cases, where an article (1) makes descriptive claims the accuracy of which is easily verifiable by any reasonable adult without specialist knowledge, and (2) makes no analytic, synthetic, interpretive, or evaluative claims, a Wikipedia article may be based entirely on primary sources (examples would include apple pie or current events), but these are exceptions.

I have a problem with this version. Although it is clear that there is a common restriction on the use of primary and secondary sources (that they be verifiable), the second paragraph places restrictions on the use of primary sources that have no parallel for secondary sources. In other words, the policy states that the use of primary sources is more restricted than the use of secondary sources. The problem is, this sentence, "However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged," puts the use of primary and secondary sources on the same level. The purpose of this sentence is to explain what "source based research is," which is fine and dandy - the sentence does not need to comment on primary or secondary sources. But by mentioning primary and secondary sources and treating them equally, this sentence can be misinterpreted to contractict the paragraph that follows. In short, it muddies the waters.

In order to clarify the waters, I have three proposals.

First, delete this (currently the second) sentence, and add in its place,

Original research that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources.

I think this is an accurate expression of our policy as it has been understood and enforced for at least the past couple of years. We discourage (nb, this does not = prohibit) the use of primary sources, and we encourage secondary sources. Slrubenstein | Talk 19:22, 11 September 2006 (UTC)

I agree. I believe that sentence used to say "research that consists of collecting ... information from existing secondary sources" (existing? as opposed to not existing?) until O^O changed it and added primary and/or secondary. I'm speaking from memory, however. SlimVirgin (talk) 19:36, 11 September 2006 (UTC)
My impression of the overall policy, considering this policy, WP:RS, WP:NPOV, and WP:V, is that editors are strongly encouraged to use secondary sources for the overall scope of articles, because good secondary sources consult a number of sources to create well-rounded information. Thus articles structured around good secondary sources are more likely to have a neutral point of view and address notable topics.--Gerry Ashton 20:53, 11 September 2006 (UTC)
I do not think it is so much that secondary sources are more likely to be NPOV (though they may be) as that when they have a POV it is very clear what the POV is (for example, in the secondary sources used in the Jesus article, critical historians, archeologists, theologians, clerics, etc.) and we can comply with our NPOV policy when using secondary sources by correctly identifying its POV. However, when we use primary sources, there is the serious risk that they be used to express the editor´s own POV which is strictly forbidden at Wikipedia. Slrubenstein | Talk 21:07, 11 September 2006 (UTC)
It is perfectly acceptable for a good primary source to address a very narrow topic and not make any attempt to place that topic in any kind of larger context. Thus primary sources are best suited for checking the accuracy of individual facts in secondary sources, or finding details that were not included in secondary sources.--Gerry Ashton 20:53, 11 September 2006 (UTC)
Yes, but this would constitute original research. You are describing precisely one of the things that scholars do when they write articles for peer-review3ed journals or for university press-published books. But Wikipedia cannot allow what peer-reviewed publications rightly encourage. Slrubenstein | Talk 21:07, 11 September 2006 (UTC)
I disagree that quoting a primary source is original research. Original research means introducing new facts or analysis. However a quote from a source is not a new fact or analysis. Wjhonson 06:02, 12 September 2006 (UTC)

Wjhonson, if you are going to participate in this discussion at least have the sense to read the proposal under discussion. Nowhere does the proposal state that quoting a primary source in and of itself is original research; on the contrary, the proposal states precisely what you are "arguing" which is that using a primary sources is not original research if it is not being used to promote a new analysis (or siynthesis, or interpretation, or explanation). Slrubenstein | Talk 16:34, 12 September 2006 (UTC)

I can check secondary sources agaist primary sources as much as I want to. If I have several secondary sources to choose from, and I find that one of them contains many discrepancies compared to primary sources, I can simply not use it in the article. Furthermore, if a secondary sources establishes that a certain topic is notable, and establishes the subtopics that should be addressed for a neutral point of view, I see no reason to not fill in some details from primary sources. I think of the secondary sources as the branches of a tree, and the primary sources as the leaves. After all, Wikipedia does not have the space limitations that the secondary sources may have had. Indeed, if I can't use primary sources for fact-checking or details, what can I use them for? --Gerry Ashton 22:45, 11 September 2006 (UTC)
Gerry, of course you personally can go through any process you want to when you decide what secondary sources to draw on. But this does not count as a criticism of my suggestion, because my suggestion above in no way would prevent you from doing that. In fact, with all due respect, i think you are missing the point. The question is not what research you do to select a secondary source for inclusion. As long as it is a verifiable source, no one can fault you for using it or quoting it. But what if I use a secondary source that you do not like - one that doesnt draw on what you consider to be all of the relevant primary sources? As long as my source is verifiable, you cannot prevent me from adding it. What you can do is insist that I properly identify the POV of the source, but that is a matter of our NPOV policy, not NOR. Now, what about your belief that my source misrepresents original sources? This is a real issue, and I encountered it in working on the Yeshu article. This is what our NOR policy - now, a month ago, a year ago - insists> you cant use your own knowledge of primary sources to fault my secondary source in a wikipedia article. If you want to fault that secondary source based on your knowledge of primary sources, you have to write a scholarly article and submit it to a peer-reviewed journal or even just a magazine, any verifiable source, and hope they publish it. But you cannot just insert the conclusions of your own original research! This is a clear violation of the policy (of course, if there is another secondary source that criticizes my secondary source, you can definitely put that in the article. But - again - what I am proposing here in NO way prevents you from doing that. Gerry, the way I see it, at least in regards to what you just wrote, you either are proposing to violate our NOR policy, or your objection simply isnt germaine to what I propose. Slrubenstein | Talk 23:21, 11 September 2006 (UTC)
Slrubenstein's passage above seems to suggest that within Wikipedia, a secondary source always trumps a primary source. I say, sometimes it does, and sometimes it doesn't. If editor A finds a secondary source that says X% of American families owned their own homes in 1900, and editor B comes along and claimes he ran the raw 1900 census returns through his computer and found the percentage wasy Y, editor B should hold is peace because that's original research. On the other hand, if editor A find a secondary source that says John Brown the abolitionist was born in 1801, and editor B comes along and says he checked the birth certificate cited by the secondary source, and it does not say 1801, it says 1800, editor B is entitled to make the correction. Editor B didn't do original research, he did fact checking. --Gerry Ashton 23:38, 11 September 2006 (UTC)
Many fields of scholarship encourage the use of primary sources, and I suspect many people reading the guidelines will encounter a statement that secondary sources are flat-out better than primary sources, and just consider the entire policy to be ridiculous, and go off and do their own thing. --Gerry Ashton 20:53, 11 September 2006 (UTC)
Anyone who wants to write an articlke for a peer-reviewed journal will certain go off and do their own thing. But Wikipedia is not for publishing original research in any "field of scholarship." It just isn´t Gerry, it sounds like you think Wikipedia ought to publish original research.
I am NOT claiming that scondary sources are "flat out better" than primary sources. I have stated something rather different: WIKIPEDIA encourages secondary over primary sources, BECAUSE Wikipedia prohibits original research. This is an inherently relative and not absolute statement. Primary sources are flat out better for some venues, just not Wikipedia. It is only in Wikipedia (according to what I propose) that secondary sources are preferred. How on earth does this say that secondary sources are "flat out" better than primary sources? Any expert scholar who construes the policy this way probably shouldn´t get involved in Wikipedia, as they would end up inserting original researsch into the articles. Slrubenstein | Talk 21:07, 11 September 2006 (UTC)
I do not advocate that Wikipedia publish original research. I just believe that Wikipedia should not write policies that appear, at first glance, as insane to those accustomed to doing research elsewhere. After Slrubenstein is done with it, the policy will have nothing good to say about primary sources, and will give the overall impression that they are reluctantly tolerated, but should be avoided as much as possible. This will come as such a shock to people used to research in other environments that they might dismiss this policy out-of-hand. Furthermore, primary sources can be better than secondary sources for fact-checking, for the simple reason that they have not been processed by as many people and there is less opportunity for typographical errors to creep in. --Gerry Ashton 22:51, 11 September 2006 (UTC)

Slrubenstein is exactly right. Wikipedia strongly prefers secondary sources in order to avoid original research. It may well be difficult for people not used to these rules, but it's critical if we're going to create a credible encyclopedia - Wikipedia is not a blog or personal website. Jayjg (talk) 21:50, 11 September 2006 (UTC)

I agree. As for Gerry's concern about "a shock to people used to research in other environments", we need to simply educate them about how Wikipedia works, and highlight the differences with other environments, rather than diminish such differences. These differences, after all, is what makes Wikipedia work. ≈ jossi ≈ t@ 23:03, 11 September 2006 (UTC)
I'm all for educating people used to doing original research about why that does not work in Wikipedia. That aside, I still have an issue with Slrubenstein, who wrote in this discussion "primary sources are flat out better for some venues, just not Wikipedia." Even within the confines of Wikipedia, there are things that good secondary sources are better for, and other things primary sources are better for. Historically, primary sources have been misused to do original research, but past misuse does not keep them from being better in certain narrow areas. --Gerry Ashton 23:14, 11 September 2006 (UTC)

So what, Gerry? You are still not objecting to anything I have proposed. Nothing I have proposed prohibits primary sources, and indeed step 3, below, explains when it is perfectly fine to use primary sources. Are there situations in Wikipedia where use of primary sources is appropriate? Of course! Not only have I never, ever argued against this, my own proposal states this. Step 3 below allows for every situation where the use of original sources is desirable and aceptable. How on earth can you claim that I am "flat out" opposed to primary sources, when I have already proposed a section that shows how they can be used without violating the policy?Slrubenstein | Talk 23:27, 11 September 2006 (UTC)

I object to any phrasing that would denigrate primary sources beyond what the current phrasing does. This is the main point that started the problem. I do not agree that wikipedia has historically disfavored primary sources in favor of secondary ones. Rather the position has been, use both, but don't use primary *predominantly*. That is what the current wording states. The alternate wording would imply that primary sources are discouraged, which is not the case in my opinion. What is discouraged is an article that is *predominantly* based on primary sources. However, an article based on say 50% primary and 50% secondary would not be frowned upon. If you agree with that last statement, then I would suggest your proposed wording be altered. Wjhonson 06:00, 12 September 2006 (UTC)

Several editors here expressly reject your interpretation of WIkipedia policy. Slrubenstein | Talk 16:35, 12 September 2006 (UTC)

And several do not, which is why we're here isn't it? Wjhonson 16:37, 12 September 2006 (UTC)
You had said "Original research that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources. " The problem with this statement is in the first word. "Original research" is prohibited period. It has nothing to do with anything else. So, this sentence draws the inference that "original research" is somehow tied to "primary sources" which it isn't. Perhaps you could rephrase it and I'd feel better about the situation. Wjhonson 06:52, 13 September 2006 (UTC)

You are right. Would just dropping the word "original" solve the problem, i.e. something like, "Wikipedia articles that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources." Or do you have an alternative? Slrubenstein | Talk 15:47, 13 September 2006 (UTC)

I would propose this wording: "Articles which draw predominantly on primary sources are generally discouraged, in favor of articles based predominantly on secondary sources". I believe that wording accurately describes that an article should be 50-percent or greater based on secondary sources. And this wording removes the inference that "original research" is somehow related to what type of source you use. Wjhonson 16:55, 13 September 2006 (UTC)

Okay, Wjhonson - I support your change completely. Now, let´s see if anyone has any objections to this version. If no one does, perhaps we have consensus on this one change? Let´s give it a day or two.

Short,sweet, and to the point. Sounds reasonable to me. Many either don't read or gloss over the section, so keeping it concise and to the point will hopefully reach the reader. GetAgrippa 12:39, 14 September 2006 (UTC)

A modified proposal

I suspect that, to be clear, we have to explain the rationale for preferring secondary sources at the same place this preference is expressed. (I would also express it as a positive "preference" rather than a negative "discourage".) Let me propose the following alternate wording for the first sentence:

Reputable secondary sources, such as authoritative textbooks and review articles, are generally preferred over primary sources in Wikipedia—primary sources often require specialists to interpret them correctly and place them in the proper context, and it is especially difficult to avoid original research when relying mostly on primary sources.

This way, we nip objections in the bud, and furthermore knowledge of the rationale makes it easier to understand the exceptional cases where a primary source might be useful. —Steven G. Johnson 00:57, 12 September 2006 (UTC)

Steven, this seems less a rationale for favoring secondary sources than a rationale for restricting the use of primary sources. How woudl you (and others!) feel about this:
Primary sources often require specialists to interpret them correctly and place them in the proper context, and it can be especially difficult to avoid brining one´s own point of view into an article when relying mostly on primary sources. Any point of view implicit in a secondary source, however, can usually be easily identified, thus complying with our neutral point of view policy. Therefore, research that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources.

Gerry, is this objectionable to you? If so, is it the spirit that is objectionable, or the wording? If the spirit is unobjectionable to you, can you suggest more precise wording? Slrubenstein | Talk 16:43, 12 September 2006 (UTC)

I prefer my formulation. The problem with primary sources is not really NPOV issues, and in any case this is WP:NOR, not WP:NPOV. I don't really understand your objection to my wording: any rationale for discouraging primary sources compared to secondary sources is, conversely, a rationale for encouraging secondary sources over primary sources. —Steven G. Johnson 18:01, 12 September 2006 (UTC)

I'm not sure if I understand exactly what Steven G. Johnson and Slrubenstein would like to take out and put in. I have created two pages, User:Gerry Ashton/sources-Johnson and User:Gerry Ashton/sources-Slrubenstein containing the Primary and secondary sources section, revised as I think these two editors would like to edit them. I invite these editors to make any necessary changes in case I have misinterpreted them. --Gerry Ashton 18:31, 12 September 2006 (UTC)

Gerry, I would rather not do this. I cannot speak for Steven Johnson, but your approach would make it seem like I "own" the proposal and am not willing to collaborate. On the contrary, you have raised two or three objections that made sense to me, and in two cases I suggested changes to accommodate you; in one case you proposed a change and I said I support it although I asked you to add a second example. This is a collaborative process and we should all seek a version that accommodates a strong majority, if this is possible. Steve Johnson suggested a change, I suggested an alternative, he rejects it. fine. But this does not mean that we now have two competing proposals. If everyone else supports Steven Johnson, I will back down. Or perhaps you ro someone else can suggest a third alternative that would appeal to more people. isn't this how a collaboration is supposed to work? Slrubenstein | Talk 19:41, 12 September 2006 (UTC)
I agree; I prefer to try to work towards a consensus rather than splintering. The reason I made an alternative proposal is that we seem to see the same objections over and over again to any statement that secondary sources are preferred. I thought it would help answer these objections, and clarify the policy, to include an immediate rationale. If other people don't agree, I'd like to hear why they think Slrubenstein's, or the current page, is better. —Steven G. Johnson 20:09, 12 September 2006 (UTC)
I think Stevenj's wording proposal is quite good. Jayjg (talk) 21:42, 12 September 2006 (UTC)
I see from Steven Johnson's edit below that I did indeed misunderstand his proposal, so I'm glad I waited before responding. --Gerry Ashton 21:57, 12 September 2006 (UTC)

A modified proposal, clarified

As for what exactly we are proposing to add/delete, let's be concrete: something like the following (old text italics, deleted text strikethrough, and inserted text bold).

Original research that creates primary sources is not allowed. However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged. All articles on Wikipedia should be based on information collected from published primary and secondary sources. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia. New sentence about preference of secondary over primary sources here.
I find this proposal acceptable. Slrubenstein | Talk 03:51, 14 September 2006 (UTC)

For the new sentence, Slrubenstein proposed Original research that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources.. I proposed Reputable secondary sources, such as authoritative textbooks and review articles, are generally preferred over primary sources in Wikipedia—primary sources often require specialists to interpret them correctly and place them in the proper context, and it is especially difficult to avoid original research when relying mostly on primary sources. for the new sentence. As I explained above, I think it is both clearer and less controversial if we explain the rationale. What do others prefer? —Steven G. Johnson 20:18, 12 September 2006 (UTC)

Stevenj, Wjhonson above proposed an alternative that I prefer. Can you and he discuss his version and yours and see what you come up with?Slrubenstein | Talk 03:51, 14 September 2006 (UTC)
You state, "primary sources often require specialists to interpret them correctly and place them in the proper context." I don't really disagree with your claim, but I do think it is a bad foundation for the policy. The consequence of this fact logically could be that expert editors be allowed to use primary sources and non-expert editors can't. Your phrasing invites this interpretation. But really, NOR should not distinguish between expert and non-expert editors and there should be one policy for everyone. The justification for restricting the use of primary sources must apply to expert editors who are capable of interpreting them properly, and not just to non-expert editors. The only justification I can think of (aside from being circular, i.e. saying usiing primary sources violates NOR because it constitutes OR, a tautology) is that even if a specialist is capable of interpreting primary sources properly in so doing s/he would be introducing his or her own interpretation (i.e. POV) into the article, thus violating NPOV. (It is not a violation of NPOV to introduce secondary sources that represent points of view, even extreme ounes: NPOV only requires that these points of view come from verifiable sources (usually secondary, as it works out) and be properly identified. Inter alia I see nothing wrong in principle with justifying an element of one policy through an element of another policy - our policies ought to be linked. Slrubenstein | Talk 21:09, 12 September 2006 (UTC)
I will not respond to Steven Johnson's question about what others prever, because I am still not certain what the exact text will be when either version is finished. As for Slrubenstein's concerns about expert interpretation introducing the expert's POV, I think the article Person having ordinary skill in the art may be relevant. There is a degree of skill usage that is ordinary, mechanical, does not introduce novel ideas or a POV, and is not original research, even though not all literate people have this degree of skill. There is a higher degree of skill usage that does introduce POV or novel ideas; it is this skill usage that constitutes original research. --Gerry Ashton 22:08, 12 September 2006 (UTC)
I prefer this version: Original research that creates primary sources is not allowed. All articles on Wikipedia should be based on information collected from published primary and secondary sources. This method of research is fundamental to writing an encyclopedia. The use of secondary sources is generally encouraged over primary sources. It is simple. I don't think a rationale is needed for the new sentence. The other proposal is good, but it is too complicated and seems a defense of policy. I see it restrictive in the "review articles" usage. Having an expertise about a subject can be problematic in writing a simple, balanced, and informative article useful for an encyclopedia. GetAgrippa 03:40, 13 September 2006 (UTC)

Wjhonson has proposed the following, which is elegant: "Articles which draw predominantly on primary sources are generally discouraged, in favor of articles based predominantly on secondary sources". As to Stevenjohnson´s version, I admire the intent but my objection stands. Could Stevenjohnson or others modify his proposed version to meet my concerns? Slrubenstein | Talk 04:06, 14 September 2006 (UTC)

Short,sweet, and to the point. Sounds reasonable to me. Many either don't read or gloss over the section, so keeping it concise and to the point will hopefully reach the reader. GetAgrippa 12:44, 14 September 2006 (UTC)
Wjhonson's short statement is fine with me. I think we'll eventually have to give a rationale somewhere, but we can get by without one for now. —Steven G. Johnson 20:09, 14 September 2006 (UTC)

An important use of primary sources

While responding to a fact template on an article I watch, Anno Domini, I noticed that a useful secondary source (Blackburn and Holford-Strevens' The Oxford Companion to the Year) alluded to the Gospel of Matthew, but did not quote sufficient material for the reader to undersand the argument. I suspect that if we looked, we could find many other examples where the author of a secondary source will expect the reader to have a certain primary source at hand in order to understand the secondary source. If a Wikipedia editor is using the secondary source, and thinks our readers may not have the primary source at hand, it would be very appropriate for the editor to quote or paraphrase the relevant parts of the primary source, so the reader can follow the arguement.

I think this is one specific use of primary sources which our guideline could not only tolerate, but encourage. --Gerry Ashton 05:25, 13 September 2006 (UTC)

Step 3: Retain our current policy concerning primary sources

This passage has been in the policy for well over a year. When I made my initial edit, I kept it; when my edit was reverted, this passage still remained. I just want to confirm that we keep it as is:

However, where an article (1) makes descriptive claims the accuracy of which is easily verifiable by any reasonable adult without specialist knowledge, and (2) makes no analytic, synthetic, interpretive, or evaluative claims, a Wikipedia article may be based entirely or primarily on primary sources (examples would include apple pie or current events). These are relatively rare exceptions and contributors drawing on primary sources shoud be exceptionally careful to comply with both conditions.

Some have criticized this as overly restrictive but after lengthy discussion I don´t think anyone could find an example of a use of a primary source in an article that violated these conditions. I think they are precise but not overly restrictive - just restrictive enough to prevent people from using Wikipedia as a vehicle for publicizing their own original research or views. Slrubenstein | Talk 19:24, 11 September 2006 (UTC)

Step 3 below only addresses the use of primary sources in articles that rely predominantly on primary sources. This leaves the question of using primary sources in articles that predominantly rely on secondary sources unaddressed. There may be cases in the latter case where including information that requires specialist knowledge to verify would be appropriate, especially if that information fills in a framework that has been established by a secondary source. For example, if a secondary source stated a mathematical problem and gave the answer, it could be appropriate to fill in the derivation of the answer from a reliable primary source. --Gerry Ashton 23:48, 11 September 2006 (UTC)

Okay. Gerry, would changing to wording (below, step 3) to, "where an article, or a section of an article," resolve your concern? I hope I understand your concern correctly. If you think this change in wording solves the problem, I have no problem making the change. Slrubenstein | Talk 23:57, 11 September 2006 (UTC)


I agree with the sentiment that an article or section based mostly on primary sources needs extra care to avoid orignial research. I would drop the part about Apple pie, since that article seems to mostly use secondary sources. Also, I think the part about what kind of articles we are talking about should come first, and the requirements second, so readers who are skimming the policy can skip part of the paragraph when it does not apply. I would write it this way:
Although most articles and sections should rely predominantly on secondary sources, there are relatively rare occasions when they may rely entirely or primarily on primary sources (for example, current events). Such an article or section should (1) only make descriptive claims the accuracy of which is easily verifiable by any reasonable adult without specialist knowledge, and (2) the editor makes no analytic, synthetic, interpretive, or evaluative claims. Contributors drawing on primary sources shoud be exceptionally careful to comply with both conditions.
The part about "without specialist knowledge" might be relaxed in articles or sections that predominantly rely on secondary sources. --Gerry Ashton 00:48, 12 September 2006 (UTC)

Gerry, I agree with you entirely. My only remaining concern is that I think if we drop the apple pie example, we need to provide another example. Can you find an article that is not current efvents, that relies predominantly on primary sources, and that conforms with this paragrasph? I think it is very important to add another example so that newbies will have a positive model as to how to comply with the policy. Slrubenstein | Talk 16:45, 12 September 2006 (UTC)

I like Gerry's paragraph (Good job Gerry), and I agree another simple example would be helpful. GetAgrippa 03:48, 13 September 2006 (UTC)
Concerning the request for an example to replace Apple pie, how about Braunfeld v. Brown, a U.S. Supreme Court decision about the sale of retail products on Sunday? According to our definition, trials and transcripts are primary sources; that might very well extend to judicial decisions. This article appears to be a straightforward summary of the case and related cases, and the article seems to stay strictly within the bounds of the primary sources.
The disadvantage of this example is it is a bit esoteric, and might not be appreciated by younger readers.--Gerry Ashton 05:14, 13 September 2006 (UTC)

I have no objection. The important thing is we have a second example. Does anyone object to this version, that Gerry has proposed and I support? Do we have general agreement to make this one change? Let´s give it a day or two to see if anyone objects. Slrubenstein | Talk 03:53, 14 September 2006 (UTC)

This is the version currently poroposed:

Although most articles and sections should rely predominantly on secondary sources, there are relatively rare occasions when they may rely entirely or primarily on primary sources (for example, current events) or Braunfeld v. Brown. Such an article or section should (1) only make descriptive claims the accuracy of which is easily verifiable by any reasonable adult without specialist knowledge, and (2) the editor makes no analytic, synthetic, interpretive, or evaluative claims. Contributors drawing on primary sources shoud be exceptionally careful to comply with both conditions.

Any objections, or are we moving towards a consensus?

I agree with the intent and content. Making it more concise would be my only comment, but due to the critical nature of the subject that may not be plausible. So all in all I agree with it and support the version. GetAgrippa 12:54, 14 September 2006 (UTC)

Step 4: Emphasize and clarify our encouragment of use of secondary sources

I propose that we then conclude with the following, which concludes with the sentence on "source based" research.

Research that consists of collecting and organizing information from secondary sources is, of course, strongly encouraged. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia.

This seems pretty straightforward and innoccuous to me. But it makes sense, after clarifying the use of primary sources, to then have a couple of sentences on secondary sources. Slrubenstein | Talk 19:25, 11 September 2006 (UTC)

yes. That would work. ≈ jossi ≈ t@ 23:03, 11 September 2006 (UTC)
This statement falsely assumes that original research and source based research are mutually exclusive. In fact, when original research is performed using primary and/or secondary sources, to the exclusion of experiments and original observations, it is source-based research. If the sentence is to be included at all, it should avoid the phrase "source-based research", because source-based research is a superset of what we want in Wikipedia. --Gerry Ashton 23:05, 11 September 2006 (UTC)

I would have no objection to deleting it is "source-based research" from this sentence, if that would satisfy Gerry Ashton Slrubenstein | Talk 23:29, 11 September 2006 (UTC)

I would like to hear what Gerry proposes would clarify matters without the possible confusion? In other words, how should we describe it rather than with the phrase "source-based research"? KillerChihuahua?!? 19:01, 12 September 2006 (UTC)
I cannot go along with this modification. Essentially Slrubenstein, wants to revert all the changes back to his original proposal, which started the whole long debate in the first place. Wjhonson 16:39, 12 September 2006 (UTC)

Wjhonson, Why can´t you go along with this modification? it is not a persuasive argument (and also not a sign of good faith) for you to disagree with my proposal simply because you disagree with my proposal. What (for both versions of this statement, the one that includes "it is 'source-based research'" and the one that does not include it) do you actually object to, and why? Provide an actual reason, otherwise you are just being argumentative and obstructive. Slrubenstein | Talk 16:49, 12 September 2006 (UTC)

My objection has been the same since day one. You are removing "primary" from this sentence. Primary belongs in this sentence. Your method of arguing, by insinuating bad-faith and obstructionist tactics, is really beneath you. I wish you would focus on the issues, instead of personal attacks. Wjhonson 17:25, 12 September 2006 (UTC)
Talking to yourself? The only personal attack I see here is made by you, Wjohnson. KillerChihuahua?!? 19:01, 12 September 2006 (UTC)
Primary doesn't belong, since Wikipedia prefers the use of secondary sources, in order to avoid original research. Jayjg (talk) 21:47, 12 September 2006 (UTC)
I find ending it with: Research that consists of collecting and organizing information from secondary sources is, of course, strongly encouraged. NPOV should be a guiding strategy in this process. This is not "original research" and is a basic approach in writing an encyclopedia.GetAgrippa 04:07, 13 September 2006 (UTC) I may be confused with all the edits. Can we compile the most accepted changes into a section for review? GetAgrippa 04:11, 13 September 2006 (UTC)
I just discovered this off-my-record discussion-page, see my remark on the Talk page. ::::: But I can agree with this last proposed phrasing. Thus I'll report this proposal on the Talk page.
I remind you once more of the disputed phrasing:
Research that draws predominantly on primary sources is generally discouraged, in favor of research based on secondary sources.
I explained on the Talk page in what way it may conflict with WP:V as well as WP:NPOV. Insofar as it does not conflict with those policies, it is superfluous.
Harald88 19:17, 21 September 2006 (UTC)

Should we archive this page?

Just what the heading says. Any objections? Anyone have any ideas as to the most appropriate way to do this? Slrubenstein | Talk 10:46, 19 September 2006 (UTC)

Yes. It is time to achive it. By any means necessary :) WAS 4.250 11:06, 19 September 2006 (UTC)

Yep. FeloniousMonk 21:17, 21 September 2006 (UTC)