Wikipedia talk:No original research/Archive 24

This is an archive of past discussions. 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 20

←

Archive 22

Archive 23

Archive 24

Archive 25

Archive 26

→

Archive 30

If I write a statement "The Pope has white hair" would this be original research? (assuming I could find no academic sources saying this.)

What I'm really interested in is how obvious something has to be before you can say it, without slapping a {{

18:31, 22 September 2007 (UTC)

Can you provide a more realistic example? This one is so obvious as to be begging the question. Vassyana 18:43, 22 September 2007 (UTC)

According to this policy, this statement could be easily slapped with the citation needed template. Many others could easily contend that the Pope's hair is not white, but a shade of grey. If you could find any source, even the Vatican web site where it said "the white-haired Pope" or whatever, then there is at least some reference to it somewhere, and the point being made is not likely to be very contentious at that point. However, if you said in an article "The youngest Pope ever was Pope XYZ at the age of 45, and his hair was already white", and then further stated "The next youngest Pope was 15 years older with white hair" (and both of those were properly referenced), you still couln't say "All Pope's have had white hair" without that statement being properly referenced. It states a conclusion that needs it's own reference. wbfergus ^Talk 18:51, 22 September 2007 (UTC)

Agreed with wbfergus in general.

WP:NOR. ... Kenosis

19:02, 22 September 2007 (UTC)

Can we avoid splitting hairs, please? ≈ jossi ≈ (talk) 19:07, 22 September 2007 (UTC)

The absence of consensus for PSTS reflected several known problems.

One was defining the categories, and there was broad agreement that this was a problem, as well as various ad-hoc fixes. I pointed out above that many ancient histories derive from even older histories. If we interpret PSTS literally, most ancient histories are secondary sources and most scientific papers are primary sources.

Another involved primary sources, even first-party primary sources, which include their own interpretive claims. These should, of course, be used very carefully. Suppose we have one secondary source which states "all X believe Y," we have several secondary sources which state "A was an X" and we have a primary source by A which states "I believe ~Y and I do not believe Y." In this case, as I've argued before, it shouldn't be OR to state that at least one X did not believe Y. (It may be undue weight, but even then noting some instead of all would be appropriate). To my mind, second-guessing the sources by stating that A was not an X, or that A really believed Y, is much more likely to involve OR. Jacob Haller 00:43, 23 September 2007 (UTC)

Simply as a point of interest and definitely NOT as an argument for or against the changes that we are contemplating... I checked out the Wikipedia articles on Primary source, Secondary source and Tertiary source. They primarily talk about the way historians use these terms... and what struck me was that the way they are defined in those aricles matched the way they are currently defined here. This leads me to conclude that the sources section was definitely written by historians. Those of you from other disciplines may want to edit those articles to reflect how you use the terms. Blueboar 13:50, 17 September 2007 (UTC)

I am not a historian and I re-wrote the sources section based on my experiences in the social science ... and I admit that they would be improved with edits from people working in the physical and life sciences. Slrubenstein | Talk 14:00, 17 September 2007 (UTC)

I will also point people to the article on Source text... perhaps this will aid in our discussions? Blueboar 14:11, 17 September 2007 (UTC)

Primary sources in history often include interpretations. Jacob Haller 16:18, 17 September 2007 (UTC)

Sure, what you are saying is (to use the above lingo) is, secondary sources can become primary sources. For example, parts of the Hebrew Bible (much of Genesis, Kings and Chronicles, for example) were probably originally secondary sources. But thousands of years later, the sources that they themselves were interpreting are often unavailable, and the context has changed, and now historians have the task of interpreting the Bible. Shakespeare may have writen things that were immediately clear to his audience when he his plays were first performed in England, but today require interpretation and explanation. I get your point, do you get mine? If we both understand one another maybe we can work together to clearer and more cuseful language? Slrubenstein | Talk 17:15, 17 September 2007 (UTC)

Diaries, laboratory notes, interviews, tabulated results of questionnaires etc. also often contain interpretations. The policy should not sound as if it's excluding the use of such interpretations. I think it needs some editing to make that clearer. (Some such might be excluded by Verifiability as not a sufficiently reliable source for that type of statement, but not by NOR.)

An (unrelated) suggested edit: change "Some secondary materials, for example many scientific publications, often include original data and are thus also primary materials." to something like "Some sources, for example many scientific publications, often include original data and thus include both primary and secondary materials." Otherwise, you're just using the word "material" to mean "source" and not getting anywhere. However, these suggestion should not hold up page unprotection. --Coppertwig 16:43, 17 September 2007 (UTC)

Do you want to go ahead and make the change to the proposed draft? Go ahead Slrubenstein | Talk 17:15, 17 September 2007 (UTC)

I think this is more than just a point of interest, it is a real problem, probably a fatal flaw, to use an article to define a policy term. It is a problem for the article, because editors are going to defend that page like they defend a policy page, which is not good for an article that may need to grow to include different perspectives as pointed out above. And it is also a problem for the policy page, because as the article changes over time, it will inadvertently change the policy description. Each page should be an expression of the consensus there, and changing one should not change the other. People have pointed out the problem with conflicting definitions in different policy pages, but taking the definitions from an article is not the solution! Dhaluza 11:29, 23 September 2007 (UTC)

We appear to agree to not use the word "source". Can we also agree to not use the words "primary" and "secondary"? People here admit that wikipedia is making up its own definition and I contend that even with a definition we make up, there is confusion and misunderstanding over the use of these words. Further, the words have no inherent meaning that makes clear what we are trying to say, as it is the use and not the material that being distinguished, and all material is primary if used as evidence of what it itself says and as secondary if used as evidence of the processes or facts that caused its existence. Whatever you wish to say about Wikipedia's OR policy, if you say it without using the words "primary" and "secondary" you will be communicating better. Please give it a try. WAS 4.250 20:27, 17 September 2007 (UTC)

I suggest "interpretive material" for the latter. Jacob Haller 21:21, 17 September 2007 (UTC)

I agree--the issue is not whether the noun is source, material, or even stuff. The issue is the use of the adjectives primary and secondary are confusing when applied to those nouns, because there is no agreement on their association in different academic fields, much less general use, so how can they be used to describe how to write a general purpose encyclopedia anyone can edit? Dhaluza 11:39, 23 September 2007 (UTC)

Point 1

Recent history that may have been missed is that there has been problem with exactly what the definition of primary and secondary sources actually are, especially when the language still strongly favoured secondary sources (as it did until quite recently, I don't know from when, but definitely a number of months ago). One of the reasons to remove the distinction is that there have been problems with the definition, and the distinction is not essential to identifying, describing, or prohibiting original research.

It should also be noted that, however long something has been a certain way, that age or stability does not automatically make it either good or correct. I'm sure there are articles on wikipedia with factual errors that have been stable (in terms of the error) for quite some time. The fact that a policy gets more attention is offset by the inertia in editing policy.

Problems were raised, challenged, demonstrated, established to the satisfaction of those involved at the time, some of whom were very skeptical. Asking the parties to demonstrate and establish them again begins to feel like obstructionism, although I am sure that that is not the intent. Can good faith and, perhaps, competence be assumed? The question to address then becomes whether the new suggestions provide for a correct, consensus-bearing, current-practice-describing NOR policy. If a particular proposal doesn't, then address how it doesn't specifically and help people to make it do so, without saying "I don't believe you about the problem that led to this".

Assume good faith that the editors who've worked on this aren't somehow trying to allow original research. This isn't a fight or a battle, it's a discussion to reach a consensus solution to certain problems and improve the policies that help us to write a good encyclopedia. SamBC(talk) 18:17, 21 September 2007 (UTC)

Point 2

It seems that a lot of the problems with the proposed removal of PSTS info aren't entirely adamant that it must be in NOR, but that it must be somewhere. Can people accept that there may be somewhere more suitable for it? The distinction between these types of sources, and between involved-party and third-party sources, and so on and so forth, touch all three of the major core policies (

here), so why not locate the definitions centrally and refer to them from each policy. If the reason for each policy to refer to them is a matter of guidance, rather than of rules, then why not put that guidance centrally as well? Those two suggestions are seperate - to me, the first (centralise definitions) seems straightforwardly sensible, while the latter (centralise the classification-related guidance) is a vague thought in my mind.

There is what's now a rather poor-shape essay at Wikipedia:Classification of sources that was actually created originally as a proposal for roughly this purpose. How would people feel about going back to this and working on it in concert with the proposal to change NOR? Those who've raised these objections would be more than welcome to join in and make sure that it says what needs to be said. If that's in place as a related proposal, it may well smooth the proposals here. SamBC(talk) 18:17, 21 September 2007 (UTC)

You may be right with "so why not locate the definitions centrally and refer to them from each policy" but this steps straight into the [[WP:ATT] debate. Also this debate is touching on the other perennial should

Philip Baird Shearer

21:13, 23 September 2007 (UTC)

Point 3

The "new opposition" referred to at the top of this section is not what's "new" at the present. What's "new" here are these proposals to significantly change

18:57, 21 September 2007 (UTC)

Kenosis, I must say that I strongly object to your language and choice of words. I know nothing about you or your history on here at Wikipedia, but even I know the Wikipedia behavioral guideline of Wikipedia:Assume good faith. I must state for the record that I strongly object to your statements, which several people have pointed out to as being incorrect, yet you continue to re-state them. This discussion has been going on for around a month, and somewhere around 20-30 different editors involved in it at various stages. If you would bother checking your facts first, instead of just spewing allegations, you would readily see this. You have been provided with several links and also asked to review the Archives, which obviously you have not bothered to do. So, either please stop making baseless accusations, or please refrain from further disputive posts. Thank you. wbfergus ^Talk 19:08, 21 September 2007 (UTC)

If we need to go over the comments I've made thus far, one by one, what they were responding to, and the responses and re-responses thereto, we shall. Now, what precisely is your objection to my "language and choice of words"? Without putting words or attributing implications that are not contained in my statements, please. ... Kenosis 19:29, 21 September 2007 (UTC)

I will also point out that the page was not locked over this proposal... it was locked over disagreements relating to the EXISTING language. This proposal was developed as a way to solve those disagreements (and indeed it seems to have done so) after the page was locked. so no... that there is resistance for such significant change is NOT evidenced by the fact that the page needed to be locked. Now, I don't think anyone is suggesting rushing off and making the change once the page is unlocked. You are correct in stating that more community input is needed. That is what we are currently doing. Blueboar 19:27, 21 September 2007 (UTC)

OK, I further reviewed the recent history, and see the distinction between COGDEN's points of advocacy, Vassyana's ideas and proposal, and the more recent new proposal developed and discussed among Bluebear, Vassyana, wbfergus and SamBC within the last two to three weeks or so. Thanks for pointing out the difference. Either way, a complete replacement of the longstanding section on Primary, secondary and tertiary sources is a significant change to a policy page. I trust that once complaints about opposition are put aside and accepted as a reasonable added participation and expansion of discussion, the discussion can continue on the substance of the proposal. ... Kenosis 19:50, 21 September 2007 (UTC)

My comment above about your comment was speciafically the part that said "developed within the past month or so by three or four users on this one talk page". Without going back over this entire I can't recollect if there were other comments by you or somebody else that were accusatory in nature or not, but I know that several comments on here in a simalr vein really got my blood boiling. We are attempting to make the policy better, easier to understand and implement across the other existing ploicies, and trying to be as above board as possible in the process. We have only reached this point with a lot of arguning, giving and taking amongst 20 or more different editors, and we are still a ways from concensus. wbfergus ^Talk 19:49, 21 September 2007 (UTC)

My points above are in no way critical of the opposition, but the point is that it is opposition to the proposal, and it's newly appeared. If you read the actual points I made, they were an attempt to explain the position leading to the proposal, explore what the nature of the opposition is (as I assume that it isn't knee-jerk, assuming good faith), and seek to address it (as in work with it and develop a way forward the addresses the issues that led to the proposal whilst eliminating the reason for objection). SamBC(talk) 20:28, 21 September 2007 (UTC)

jossi and Odd Nature, please explain in detail what you object to in the proposal, perhaps giving examples. Philip Baird Shearer said that we can summarize a tertiary source but we can't summarize a primary source such as a treaty. That may be something that's said by the source-typing policy but not by the new proposal. I did some thinking about that. (As an aside, I'm not sure we absolutely can't summarize a treaty. I think we may be able to say things like "this section talks about genocide", if that word appears, or "The treaty consists of 9 sections which are titled ...") I think it's not the type of the source that prevents us from summarizing it; rather it's the amount of material that comments on the source. The more reliable-source material there is (or is likely to be in future, for example for a treaty that's just been written) that interprets and comments on a source, the less we can summarize it ourselves as opposed to quoting directly. This could apply even to an encyclopedia article. The wording of a signed article could become controversial as evidence of the state of mind of the author if the author is charged with a crime, for example. It would then become un-Wikipedian to summarize what it says ourselves rather than relying on commentary about it. Quoting would still be fine. So, maybe some tweaking of the wording of the proposal is needed. --Coppertwig 22:58, 21 September 2007 (UTC)

Actually, I think summarizing a treaty is covered under the proposal... If an editor summarizes a treaty, that would be a statement of interpretation/analysis/conclusion etc. ... thus, for it not to be OR, he/she should cite to a reliable source that contains the summary and not to the treaty itself. If the editor stuck to the source - ie quoted the exact wording in the treaty(preferred) or closely and accurately paraphrased it, then he/she could cite the treaty itself. Blueboar 23:30, 21 September 2007 (UTC)

Novels are primary sources. According to Blueboar's reasoning, editors must not summarize novels; they must find a summary in a published review or similar secondary source. Which of course is an absurd position. --Gerry Ashton 23:40, 21 September 2007 (UTC)

I'm not entirely sure that that interpretation is correct, but I'm also not sure that the current policy is entirely comfortable with summarising a novel without reference to a source for the summarisation. Summarisation, especially of fiction, is notoriously subject to spin and interpretation, and that's original research. Summaries can appear, but we shouldn't be doing the research to write them. How about this as an indicator of OR - if people can disagree over it without anyone ever referring to sources, it's likely to be OR. SamBC(talk) 00:16, 22 September 2007 (UTC)

Good source-based non-original research requires skill, some understanding of the subject matter (after the research is completed), judgement, and decision making. Summarizing material is a basic skill of source-based research; just because it can be done badly does not mean source-based research can do without it. After all, we're talking about research, not monkeys and typewriters. --Gerry Ashton 00:30, 22 September 2007 (UTC)

Bear with me, I'm trying to find the right way(s) to say this. Summaries can be written that do not involve interpretation/analysis/conclusion etc, in the sense that leads to OR. They can be written in a way that does. Summaries are not "right out" under any version of the policy that's been discussed (that I'm aware of), including the current version. It takes care to write a summary that's not OR, and it also takes care to check whether a summary does introduce OR (in the form of interpretation/analysis/mongooses/etc). The PSTS nature of the source isn't actually entirely relevant to that, which is why it's been suggested that PSTS isn't a vital and inherent aspect of the NOR policy. All that said, you raise an interesting point about summaries generally, and it might be helpful for the policy to have a section (seperate to that under discussion) discussing summarisation. SamBC(talk) 00:47, 22 September 2007 (UTC) Please note that a word in the preceding post has been replaced with an irrelevant word as a vent to sanity

Churchill once said of Chamberlain "he viewed world affairs through the wrong end of a municipal drain-pipe." I think that these debates are very constructive because we all tend to edit in certain areas and the effects of these sections in policies vary on different types of articles. These conversations improve our understanding of the issues by presenting the information in a broader context and it is often the unforeseen consequences of these policies and changes to them that cause the most problems for people.

BTW, An example of where a primary source is also a very good secondary source is the

21:49, 23 September 2007 (UTC)

The definitions of PSTS are not critical to NOR because they are not used outside of the PSTS section. I have rewritten that section without using the defined terms: primary, secondary, and tertiary:

Use of sources

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. However, care should be taken not to "go beyond" the sources or use them in novel ways.

• Factual sources - Facts that have been published by a reliable source may be used in Wikipedia, but only with care, because it is easy to misuse factual sources. For that reason, anyone—without specialist knowledge—who reads the factual source should be able to verify that the related Wikipedia passage is true to the factual source. Any interpretation of factual source material requires a reliable interpretive source. Examples of factual sources include archaeological artifacts; photographs; historical documents such as diaries, census results, video or transcripts of surveillance, public hearings, trials, or interviews; tabulated results of surveys or questionnaires; written or recorded notes of laboratory and field experiments or observations; and artistic and fictional works such as poems, scripts, screenplays, novels, motion pictures, videos, and television programs.

• Interpretive sources draw on factual sources to make generalizations or interpretive, analytical, or synthetic claims. A journalist's story about a traffic accident or a Security Council resolution is a interpretive source, assuming the journalist was not personally involved in either. An historian's interpretation of the decline of the Roman Empire, or analysis of the historical Jesus, is a interpretive source. Wikipedia articles should rely on reliable, published factual and interpretive sources used appropriately.

• Some publications such as encyclopedias that sum up other sources are more reliable than others; within those sources, some articles may be more reliable than others. For example, articles signed by experts in
  Encyclopaedia Britannica
  , and encyclopedias of similar quality, can be regarded as reliable interpretive sources. Unsigned articles may be less reliable, but they may be used as factual sources so long as the encyclopedia is a high quality one.

An article or section of an article that relies on a factual source should (1) only make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge, and (2) make no analytic, synthetic, interpretive, explanatory, or evaluative claims. Contributors drawing on factual sources should be careful to comply with both conditions.

Comments

I tried not to change too much, to remain as true as possible to the original. But once the sacred cows are banished, I think it shows this is not a good example of clarity. It needs lots more work, but we can't get that done until we get over the religious arguments over PSTS, and actually get down to the real work of editing. Dhaluza 01:30, 23 September 2007 (UTC)

What is a factual source? Are we now reinventing distinctions? ≈ jossi ≈ (talk) 01:37, 23 September 2007 (UTC)

Yes, in fact, I was trying to show an alternative to the confusing primary/secondary distinctions using plain English. I think if you ask a friend or relative who has never seen Wikipedia "what is a factual source?" and "what is a primary source?", you will probably get a more reasonable answer to the former than the latter. Dhaluza 11:21, 23 September 2007 (UTC)

Just to pose a very simple example, illustrative of numerous real situations that have already occurred on the wiki: Take Aristotle's Metaphysics and the various interpretations thereof. A lot of "know-it-all's" in the world like to quote Aristotle and are frequently at odds with the range of scholoarly interpretations of Aristotle's voluminous works. Where does Aristotle's Metaphysics fit into this proposed schema? Is it a "factual source"? Or is it an "interpretive source"?. If Metaphysics is itself an interpretive source, why can't we just interpret the interpretations. Isn't it better to go right to the "horse's mouth" rather than some Ivory Tower interpretationist? .. Kenosis 01:53, 23 September 2007 (UTC)

I could give many other examples myself. It strikes me that these categories may make sense for journalism and the like, but I'm not sure they make sense for many areas of knowledge. While we often think of philosophy and religion as being primarily interpretive, this is equally true for science as well. Most scientific theories are efforts to interpret the available evidence, and the facts on which scientific theories are based -- data, collections of measurements and the like -- are generally not appropriate for encyclopedias. It strikes me that this proposal comes close to imposing an

reliable and whether that source said something, and tell whether the artcle content sticks to the sources or draws inferences beyond them. There is simply no need to get trapped in the hopeless swamp that attempts to classify the complexity of human knowledge into these types of epistemological bins tends to lead to. It's a swamp we can and should avoid. Best, --Shirahadasha

02:28, 23 September 2007 (UTC)

First, I am by no means promoting the use of "factual" or "interpretive" as being good (or bad) descriptions, just plain English alternatives to the obscure and arbitrary "primary" and "secondary". This was not an activist attempt to impose epistemology (if anything, clinging to PSTS is imposing agnotology, but we are digressing from the point, which is using plain English). I don't think this example is a suitable alternative explanation without further development. As I said, I tried to stick to the original text, and just substitute terms, and found this revealed basic structural problems with the existing descriptions. I think the comments above also show that the PSTS text does not really explain NOR without itself requiring OR.

In the transition, I had to drop the primary source examples of eyewitness accounts because these require more careful explanation--for example is the witness an expert eyewitness or not? We can't use eyewitness accounts to draw conclusions, but we can use them to highlight differences, etc. The other example I though of that could be further developed is the

Rosetta stone. As a factual source, you could say it was a stone tablet, but that's about it. Any further analysis would require expert interpretive sources. The issue is that we do not need to try to sort sources and stick them in a particular bin, we need to carefully describe what is and is not appropriate use of sources on Wikipedia. Dhaluza

11:21, 23 September 2007 (UTC)

---

Here is another example, dropping classification of sources altogether, focusing on use only. I've dropped the examples for clarity of the example, but appropriate explanatory examples probably should be added to each section.

Use of sources

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. However, care should be taken not to "go beyond" the sources or use them in novel ways. This policy places certain restrictions on the use of sources to preclude editors from including original research in Wikipedia articles.

Facts

Wikipedia does not bear witness to any fact. All statements of fact in articles must be supported by a previously published reliable source. Factual statements must be made with caution, because misstatement of facts is a common problem in Wikipedia articles. The facts should be presented so anyone—without specialist knowledge—who reads the cited source can verify that the related Wikipedia passage agrees with the source. Any interpretation of facts or evidence requires a reliable source for that interpretation. Wikipedia editors must be careful to not make statements of fact based on their own observations or evidence, unless they are also supported by reliable published sources.

An article or section of an article that makes a statement of fact should (1) only make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge, and (2) make no analytic, synthetic, interpretive, explanatory, or evaluative claims. Contributors making statements of fact should be careful to comply with both conditions.

Interpretation

Wikipedia does not publish original thought. Using raw evidence or statements of fact to make generalizations or interpretive, analytical, or synthetic claims is original thought. Wikipedia editors must be careful not to include their original thoughts in Wikipedia articles, but must instead depend on reliable published sources to make these interpretive claims.

Some publications such as encyclopedias that sum up other sources are more reliable than others, and within those sources, some articles may be more reliable than others. For example, articles signed by experts in

Encyclopaedia Britannica

and encyclopedias of similar quality can be used for interpretation of facts and evidence. Unsigned articles may be less reliable, but they may be also be used for interpretation so long as the encyclopedia is a high quality one.

Comments on this example

This is an example of how to describe NOR without resorting to the unnecessary distraction of source typing. It focuses specifically on the use of the sources, not the nature of those sources. It basically says the same things as PSTS (excluding the examples) without using the adjectives primary or secondary, or other replacements. It may not cover every point in this iteration, but could with minor revision. I think this is a much simpler, more understandable, and less confusing explanation of sourcing as it relates specifically to NOR.

Step-by-step

Looks like I'm one of the first to wake up from a few hours of sleep. I'm starting the day by browing through the history pages of NOR. One of the first things I've noticed, is that any change to the policy has been met with fierce opposition, even changes that corrected obvious inaccuracies. One example is [[1]]. This was a fairly obvious and simple correction, changing the policy wording from saying there were three other main content policies but only listing NPOV and V, so the editor changed it to simply say "others" instead of "three others". This was immediately reverted back to "The other three are Wikipedia:Neutral point of view (NPOV) and Wikipedia:Verifiability (V)." with the revert labeled as "Restore longstanding version, no discussion, no consensus for this"., basically accusing the first editor of making a change without any support. This was a rather childish example, the editor made a simple minor correction in good faith that corrected a glaringly obvious mistake, but it took another round of reverts to get paast just that one simple case. The edit comment was glaringly false as well, as a simple edit history check shows that typo didn't appear until earlier that same day (two hours earlier). That is hardly "longstanding".

I next noticed, spurred in part by a previous post of

until June 22, 2007, so claims that this (PSTS) is long-standing policy are untrue. Before the edit by SlimVirgin on Oct 23rd, the "Sources" section only talked about primary and secondary sources, tertiary sources weren't mentioned. Just since Oct 23, 2006, there appears to have been almost 500 edits to this policy, though some (I have no idea how many) were vandalism and the subsequent reverts. So stating that this policy has been stable seem blatantly untrue as well. Simply going to the history and setting it to show 500 edits at a time, it is still less than a year ago at the previous 500th edit, so 500 edits in less than a year (about one and half edits per day to a policy), is not very stable, at least within the last year.

With these brief observations out of the way (only included to disprove statements of "stable" and "long-standing"), lets move on to what Wikipedia is at it's very basic core. I am not talking philosophically of what Wikipedia should be, or strive to be. I am talking that at it's very base, Wikipedia is simply a database, though one with a very good user interface (though a few small improvements could be made, more later).

Now, as anybody with more than just a tad more than rudimentary knowledge of databases knows, proper database design tries to limit the amount of redundant information and therefore redundant coding, for consistency and simplicity. This is not meant as derogatory in any way, just a statement of fact. I work everyday with many smart scientists that even after 12 years of usage, still can't grasp the concepts of proper database design. Anyway, let me try to go back to the insurance analogy.

Insurance companies write hundreds of thousands or more insurance polices each year. These are not all the same, but simply boiler-plated together for the customer. For example, a car policy, a business policy and a house policy will all contain a clause about "Acts of God". This clause, specifically the definition, is not included and defined in each type of policy. There is simple code in the program that tells it to also include the standard definition of "Acts of God" in each policy. The definition itself only exists once within the database, but that common definition is used by multiple "things" (policies). Databases use the same concept, primarily with what are sometimes called "List of values" (or select lists), though there are many more common usages as well. What this means, is that instead of defining something over and over and over, you simply define it once and then use that one definition over and over. In the long run, this makes maintenance much easier. When the definition changes, you make the one change in one place, and instantly everything that links to it begins to use the same standard definition again. You are not forced to search through the entire database for every occurrance to make the same change over and over again. So if some lawyer comes along to the insurance company and says "We need to add this extra definition of an act of God", they pull up that definition, make the change, and 'poof', all new policies instantly begin using the new definition. They do not have to go through all policies trying to see if that definition exists there and then changing it.

This is all we are proposing for this policy. Hopefully

could see after implementation that it is easier and more consistent. However, before we can get there, there's an awful lot of work that needs to be done first, and we also have to take into account the concerns made by other editors, even if they do "over-exaggerate" some claims. So, trying to toss out ideas on how this 'could' work (to even stand the proverbial snowballs chance), this is what I see we need to do:

• Take all of the existing definitions of "Source" type issues from the existing four policies, and try to combine them into not only one location, but also in a coherent form. This in itself is a formidable task. One or two people could probably get the bulk of it accomplished, but then editors personally involved in the other related polices would also need to be involved to help tweak it per their respective policy backgrounds, ensuring that no salient points are missed or otherwise weakened. This may also require some 'policy' specific sections, I don't know. It seems like including all of this in
  WP:RS
  is the most logical place, but that current guideline also needs improvement, as Wikidemo and jossi can attest (I believe they are both currently working on that page).
• After this phase is accomplished, then that page (whatever and wherever it is), would need to be presented to the entire Wiki community for comments, to see if there would be consensus for making it a policy. Now just that alone would make most people automatically reject it. This would also have to be presented with the additional explanation that if this was accepted, then the applicable sections from the other policies would be redundant and could be removed from those policies, to be replaced and (two-way) linked appropriately. This would result in five policies instead of the current four, but would result in more easily maintainable, consistant, and understandable policies. Now this would be far easier if Wikipedia had something like an #include directive, like several programming languages do. Then, each policy could simply have the #include statement to pull the appropriate section from the 'sources policy'. Instead though, I think we'd have to jump through the hoops of just using links back and forth.
• If we can achieve the creation and acceptance of a 'sources policy', then changes to this and the other policies could occur, as at that point, removing things like PSTS would be redundant, and with a link to the new policy, nothing would be lost.

I'm may be missing something else up above, but that is how I see we have any chance of ever improving this. If there are people who will revert even obvious typographical errors just because it's a change or over-exaggerate their claims so others who don't bother to research the history will more easily fall in agreement behind them, then we need to come up with a step-by-step approach that clearly shows the benefits. About the only way to actually accomplish this, since no amount of explanations seem to get through to some people, will be with viable alternatives that they can easily see. To be fair though, it does help assuage their fears that won't remove something like PSTS now, and it's replacement home will be worked on. We all know full well how long that usually takes (years), and there has to be an established place for it, probably at a policy level. Including these in NPOV, V or BLP doesn't accomplish much even though those are already policies, as those really aren't logical places for this 'subject'. Either a major rewrite of RS, or the creation of something similar seems to be the only way, though I could be wrong.

Accomplishing this would take probably 3-6 months, depending on how long it takes to create the new 'centralized' home, and then addressing concerns brought up by other editors and the community in general. I really don't see how it could be accomplished any faster, but I suppose it could. Like some changes to this policy (and I'm sure others as well), there could easily be times when the bulk of 'objectors' are absent from Wikipedia or otherwise busy for awhile, and a change can be made and be present before most know what happened. But that is real back-door and I wouldn't want to be a party to it. I honestly think these proposed changes would be a great benefit to Wikipedia, making 'policies' easier to understand (especially for the newcomers) and more consistent in enforcement of policies. But, we also need to do it 'above board' and with viable alternatives availabable at each point in time or it will surely fail. wbfergus ^Talk 11:32, 23 September 2007 (UTC)

Your example of ridiculous reversion is right on target. Some editors believe every change to policy must be discussed to death, and any change should be reverted "on sight" if it hasn't. This is not consistent with

WP:POLICY

but it is an all too common bad-practice. Only things that are new or controversial need to be discussed. Simple text edits can be discussed with edit summaries, and should not be reverted by an editor unless they are 100% opposed to them--even if it is 90% disputed, the undisputed 10% should be left in so progress can be made. It's the same as a disputed bill--you pay the undisputed amount.

I also agree that PSTS is only half-baked, and the cost/benefit to NOR is not clearly favorable. Having a widely applicable unambiguous set of definitions for types of sources may be useful if it existed, but it clearly does not exist now, and is not essential to NOR. So the sooner we jettison the excess baggage of PSTS from NOR and define OR in plain English, the better. Those who are concerned about coming up with source type definitions can shepard PSTS through the process of redefining it and possibly re-incorporating it later if a consensus forms around it. But there is clearly no consensus for it now, and keeping the page protected to protect PSTS must stop as soon as possible.

Your insurance policy example is somewhat consistent with my experience in writing technical specifications. In a spec document, you must be careful to define a term only once, and use that definition consistently throughout. If a term is defined multiple times, you don't get the sum of the definitions, you only get the least common denominator. Legal documents like specifications, contracts, and insurance policies tend to favor defining terms in the document. But general purpose documents like instruction manuals should avoid using defined terms as much as possible, because in a non-legal setting, it leads to confusion. Since we depreciate wiki-lawyering, I would suggest that we should write policy more like a manual than a spec, and also avoid defined terms as much as possible. That is why I have been arguing for dropping PSTS and explaining NOR in plain English. Dhaluza 12:12, 23 September 2007 (UTC)

I'm not in the insurance industry, so I can't accurately state exactly how the process works, I simply used that as an example that I think most people can easily comprehend, based upon my last 27 years experience working with databases, both design and programming. I also agree with your point about when there are multiple defintions (as is the norm with Wikipadeia policies currently, since they are to be used in conjunction with each other) that one side takes the least most denominator (the most restrictive) but there is also another 'group' that will always push for the least restrictive as well, since it isn't explicity spelled out, and with conflicting definitions, what is actually meant is open to a very vigorous debate (from my experience). wbfergus ^Talk 12:25, 23 September 2007 (UTC)

With respect, you are mischaracterizing the situation. I have stated already that tertiary sources as a concept has dropped in and out of this policy for most of its existence. When i have stated that it has been fairly stable, I mean the idea of original research and the distinction between primary and secondary sources - this is the main point under contention, the issue of tertiary sources is secondary (no one here has made it a principle point of contention). Be that as it may, let's get the story right. First, do not confuse the shortcut PSTS with the presense of the distinction in the policy itself. It is a non-sequitor to claim that because the shortcut is less than a year old, PSTS being a component of the policy itself is less than a year old. In fact, Reddi introduced the distinction between PS and T sources in this edit on February 13, 2004 - thre years and seven months ago. More to the point, that was only the twenty-third edit ever made to this policy (a point I have made already, above). So please, let's not say that the distinction is less than a year old. Now, just in case you missed it, I repeat: "tertiary sources" has dropped out as well as in various times in the history of this policy, and we have also played around with defining terms in an attempt to be consistent and to match practice. I certainly think most people who have played a role in developing this policy are open to continued discussion on its value. What people have been arguing is the importantce of primary and secondary sources, and in this regard the policy has been pretty stable. Slrubenstein | Talk 14:33, 23 September 2007 (UTC)

I don't mean to impugn the integrity of anybody, especially somebody who I may disagree with at times but have required a respect for their judgements, but it appears that you have also stated a case that the "sources" section hasn't been very stable, if teriary has dropped in and out of the definition over the years. The overall policy in general may have been relatively stable, but browsing the history shows that the "sources" section itself has undergone numerous revisions. I have another semi-comment, but I'll include it below Jossi's comment below. wbfergus ^Talk 14:50, 23 September 2007 (UTC)

Your correction to wbfergus's conclusions are noted. I actually created the

WP:NOR because that was counter-intuitive) and that should not be taken as the start date for the concept, only the shortcut. But I think it is important to also note that the diff you cite characterizes WP as a secondary source, which is not consistent with the current policy framework. So this just goes to show that the definitions of PSTS have been confusing from the start. Dhaluza

14:56, 23 September 2007 (UTC)

Perusing the archives, I went to one covering the time-frame for when PSTS actually came into being

Wikipedia_talk:No_original_research/archive15. I see extremely little discussion there about it. There was sombody who posted that another person was making lots of edits to the policy (24 different ones), but not a single edit on the talk page announcing the changes or proposing the changes.There was subsequently no additional discussion other one other editor stating "Skimming over the diffs, they all appear quite reasonable". I also don't see any calls for opinions or other types of votes on if the changes met with any concensus of those who knew that the change was being made. I also fail to see that any announcement of the change was ever posted at the Village Pump. So, it appears that this 'addition' was made without widespread community concensus, just the opinions of the few who happened to be following this page at that point in time, based upon the lack of dissent from the few that were actively involved then. Yet now trying to improve the policy for understandability and pertinence has to involve widespread community involvement. So, just a few can get additions made to a policy without widespread involvement, but fixing it after the fact requires widespread community concensus? I guess this helps to highlight why we need to make this proposed change honest and completely above-board, so in a year or so, others can't come in and claim that this was accomplished behind closed doors. wbfergus ^Talk

14:18, 23 September 2007 (UTC)

Something that remains in a policy page unchallenged for a substantial period of time, carries the implicit presumption of consensus. If that implicit consensus is challenged, yes, you will require widespread community consensus to change it. ≈ jossi ≈ (talk) 14:40, 23 September 2007 (UTC)

Saying primary sources should be rare and saying that claims based on secondary sources are presumed to not be original reseach has been contested from the very beginning. By the way, wbfergus, the Wikipedia software does support the equivelent of include. It is called transclusion and is done for every template. You can do it for nontemplate space pages also { { : name of page to be transcluded } } without the spaces. WAS 4.250 15:08, 23 September 2007 (UTC)

This is a novel interpretation of

WP:ATT), or that consensus has changed over time, then the previous consensus is only of historical interest, and is not considered a precedent. Dhaluza

15:02, 23 September 2007 (UTC)

Again, I don't mean to impugn the integrity of anybody. It just seems strange that the current proposal has been deemed to require widespread community concensus and announcements, etc., while the inclusion of of this did not. Can you please refer me to a change in overall Wikipedia policy in the last year that now requires this to be done? Or can you please show me where I missed something in this policies archives (or on the Village Pump archives) that shows that these additions were announced and widespread community involvement (concensus) was being requested? I have been unable to find anything like that so far. All I've seen (though maybe I've been looking in the wrong place), is what I have posted above. It appears (I am not saying this is the case) that only a relative few people, probably less than 15, even knew that these additions were made at the time they were being made. I fail to see how this can be construed as "implicit concensus" if nobody knew about it. Thank you for any clarification you could provide on this or other related issues.

Also, let me restate once again that I absolutely do not propose anybody trying to force this proposed change through any back-door. I personally feel that what I stated above in my observations, while actually detracting from a point I was trying to make (bad thing about editing while still sleepy), are the perfect example of why this should not happen. Everything should be completely honset and out in the open, with the widespread community asked for their opinions. Only in this manner will future editors see that the entire process was completely above board and no small group forcing through changes that affects hundereds of thousands of users. wbfergus Talk 15:04, 23 September 2007 (UTC)

W.r.t. the history of the PSTS section, the assertions above by wbfergus are incorrect, erroneous by selective omission, or at the very least very misleading. Wbfergus says he looked into the history. He didn't notice that "Primary and secondary sources" has been part of this project page for over three years? I too have had a recent opportunity to quickly review the history of both the policy page and the talk page. The primary source/secondary source distinction has been in place well in excess of three years, and arose out of direct guidance from the WP founder. Some time later it became a discrete section called "Primary and secondary sources". What happened in October of 2006 was that it was expanded to "Primary, secondary and tertiary sources". This expansion from primary/secondary to PSTS explained a great deal, and it remained extremely stable for nearly a year, during which time it has served the community quite well. This is why it required little or no explicit argumentation about the existence of

WP:PSTS on the talk page. ... Kenosis

14:54, 23 September 2007 (UTC)

Both of you are taking edits out of context.... while the policy has indeed included discussion of Primary, and Secondary sources for a long time... for most of that time the distinction was focussed on explaining what type of source Wikipedia should be... it wasn't talking about the types of sources that should be used in Wikipedia. That is a much more recent change. Blueboar 15:12, 23 September 2007 (UTC)

Okay, this is getting more convoluted than I intended it. Would the respondents of this section mind if I reformetted this, so that the point I really wanted to make with my example above is separate from the other discussions about my own personal observations? This would allow things to be grouped more accurately together, and hopefully not detract anymore from the example I was trying to provide. I feel I need to do more research into the history of this and the other policy pages, and also of Village Pump, but none of that has any bearing whatsoever on the example that I really wanted to get across. Thanks.wbfergus ^Talk 15:11, 23 September 2007 (UTC)

Dhaluza writes, "This is a novel interpretation of

WP:POLICY. A policy statement must have consensus to remain. That it remained for any period of time only means it had implied consensus in the past" and makes a classic error in confusing article pages from policy pages. The very wiki nature of this project means that articles can and should be updated and improved regularly, and consensus can change whenever a new editor brings new knowledge. But for this to work, the opposite must be the case with policy pages. Dhaluza's argument suggests that as soon as one new person comes to Wikipedia who rejects a policy, the policy no longer has consensus. That is absured for two reasons. first, imagine how easy it would be for someone who simply rejects Wikipedia policies to come, say "I do not agree, therefore there is no longer consensus" and remove the policy. It defeats the whole point of policies, which leads me to the second reason: policies are pre- or proscriptive, not descriptive, and one function is to instruct newbies as to how to contribute to Wikipedia, and also as points of reference for reverting, blocking or even community bans of editors who refuse to follow our policies. This means that policies must have more inertia than articles, and consensus must reflect the weight of time. The opposite has to be the case with articles, as with time we have more and more editors who hopefully have access to more and more research and can add more and more to articles - we actively seek improvement of articles so past versions carry weight only insofar as they comply with our policies and standards. That means we can just let people change policies carelessly. Past consensus carries a good deal of weight. It provides the model for future editors to follow, if they are to make positive contributions to the project. Slrubenstein | Talk

15:15, 23 September 2007 (UTC)

Exactly. I have seen newbies challenging WP:NOR on the basis of being original research and WP:NPOV on the basis of being not neutral... WP policies are not articles and require a certain degree of stability, in particular at this stage of the project. ≈ jossi ≈ (talk) 15:31, 23 September 2007 (UTC)

WP:Consensus addresses the extreme example you cite and attempts to strike a balance between stability and stagnation. And we are clearly now in a period of stagnation that must end as soon as possible. You cannot dismiss all proposed changes as attempts to undermine the policy. Clearly an attempt to dissolve the NOR policy would require wide discussion throughout the community. But using a different word to describe the same thing does not. Extreme examples are not needed, and not helpful. Dhaluza 15:47, 23 September 2007 (UTC)

(outdent)Okay, the

17:18, 23 September 2007 (UTC)

Motivations

It has become obvious to me that we have three competing but overlapping groups here...

• 1) Those who have problems with the language of PSTS.
• 2) Those who don't have a problem with the language of PSTS per say, but do not think it fits in the WP:NOR policy.
• 3) Those who feel that PSTS is vital to Wikipedia, and do not think it should be touched in any way shape or form.

The proposal to replace PSTS with more NOR focussed language came out of a compromise between the first two groups. Essentially the compromise says: Remove PSTS from NOR, and deal with it elsewhere. The problem is that this compromise did not take into account the third group. They need to be included in the compromise.

The third group is not going to agree to anything that demotes or weakens PSTS. At the moment they see the proposal as doing just that. The third group has to feel confident that the concepts and basic language contained in PSTS are protected before they would agree to anything.

So... I think the only solution to this three way argument is make it a complete package... we need to hammer out a proposed PSTS policy/guideline page to go along with the proposed replacement section for NOR. Blueboar 15:06, 23 September 2007 (UTC)

I agree entirely. For the record, I would consider myself part of "group 2". wbfergus ^Talk 15:12, 23 September 2007 (UTC)

I'd be happier with one tiny step at a time. First add some variation of the context/history/justification "voted" on above. Second remove claims that indicate primary sources should be rare. Third add something that notes that claims based on secondary sources are not presumed to not be original research. Does anyone have a problem with these three minor changes? WAS 4.250 15:23, 23 September 2007 (UTC)

There appears to be some obfuscation and misrepresentation going on here, possibly to satisfy some sort of narrow agenda. I suggest that what has served the community well for years, particularly years of growth and growing acceptance, should not be tossed aside by a tiny group of editors on some sort of personal mission. If there are a substantial number of editors in the 2nd and 3rd categories, then I agree; make sure that the PSTS situation is dealt with first and its continued existence in its present form or some even more explicit form guaranteed, then make whatever minor changes to format and organization that are required.--Filll 15:22, 23 September 2007 (UTC)

While your attempts at broad consensus building are commendable, I think the starting point is incorrect. We need to start from existing consensus and build up from there. The existing text is strongly defended by group 3 as you point out, but that does not mean it has consensus support. Don't take this out of context, I'm not saying the views of group 3 should be ignored, but they should not be given undue weight either. The suggestion that because this text has been around in some form for some time, and therefore must remain, is not consistent with WP:Consensus. I think we can make sure WP:NOR says everything it needs to say without using the words "primary" and "secondary" because it is not essential to WP:NOR. If it is essential to WP, then that position needs to attract consensus in another forum. For the record, I am in group 1 and 2. Dhaluza 15:39, 23 September 2007 (UTC)

That is an argument that does not have basis in reality. Wikipedia policies are official and have implicit widespread support . If you, or a small group of editors what to challenge that, the burden is on you to demonstrate lack of consensus. ≈ jossi ≈ (talk) 15:44, 23 September 2007 (UTC)

Current wording of official policy pages are a codification of current convention and common practice which already have wide consensus. Incremental improvements are always welcome, but major changes such as removal of material that has been stable for years, cannot be contemplated unless there is wide support. That is basically the issue here. ≈ jossi ≈ (talk) 15:47, 23 September 2007 (UTC)

I agree that there is a burden of proof to show the policy does not reflect consensus, but that burden was met long ago in this discussion. This section attempts to summarize the discussion comments into three camps, only one of which supports the inclusion of PSTS in NOR in its present form. Dhaluza 15:58, 23 September 2007 (UTC)

Not really. Five or ten editors' arguments does not demonstrate lack of consensus. Look at what happened with

WP:ATT: it had the active participation of hundreds of editors over a period of 6 months, it was announced at the VP, the mailing list, the WP:POST, and was still rejected as lacking consensus. ≈ jossi ≈ (talk)

16:10, 23 September 2007 (UTC)

Good example, but I disagree with your conclusion because the absolute numbers are not relevant. ATT failed because it had a narrow consensus, and failed to attract a broader consensus. The same is true for PSTS. Dhaluza 16:24, 23 September 2007 (UTC)

That is not precisely true. ATT went down primarily because Jimbo vetoed it. It had consensus until he vetoed it, then it lost consensus (evidence of the sway he has here). If we learn anything from ATT it is this: if Jimbo does not think that a policy has or ought to have concensus, he makes it very plain on the talk page of the policy/proposal. Capiche? Slrubenstein | Talk 16:58, 23 September 2007 (UTC)

That Jimbo vetoed it is one thing. This "veto", so to speak, occurred only after a month-and-a-half of wide confusion at the grass-roots local article level over the change in the layout of basic policy. People all over the wiki had come to rely on the three core content policies in their work on the wiki. Similarly here, people all over the wiki have come to rely on the primary/secondary source distinction, and more recently on the WP:PSTS distinction. There needs to be an extremely strong set of reasons and a far broader base of participation in this discussion here to even seriously consider changing this longstanding, basic policy approach to avoiding original synthesis, or original research. ... Kenosis 17:21, 23 September 2007 (UTC)

Narrow consensus, Dhaluza? 300 editors is not narrow. What is narrow is the number of editors challenging long-standing wording in this policy. ≈ jossi ≈ (talk) 17:34, 23 September 2007 (UTC)

As for the veto, I would say that is an extreme characterization. Jimbo just joined the chorus questioning the consensus, and the policy proposal fell. Actually I think it's a shame that WP:ATT could not be salvaged, as it was a good idea at bottom (I even worked on it a bit). As for narrow, 300 editors is narrow compared to the broad WP community. The true test of consensus (or not) is whether something gains or loses support as the community is broadened. I think the basic problem with the formulation of PSTS is that there is a core group of people who support it, but as you get away from that core, the support diminishes because it is confusing and not universally applicable. I agree that the discussion should be broadened, but I think that as it is broadened, and the pros and cons are weighed, the present formulation of PSTS will continue to fail to show a broad consensus support as an explanation of NOR. Dhaluza 20:02, 23 September 2007 (UTC)

I've noted above that PSTS defines most ancient histories as secondary sources or mixed primary/secondary sources since they usually include older histories and only sometimes include eyewitness accounts, since they include the historians' own interpretations, and since they don't resemble the "examples of primary sources." Nonetheless, common practice, including common Wikipedia practice, regards these as primary sources, or groups ancient sources into their own category. (P.S. I think we can agree that the literal reading of PSTS which makes these secondary sources is not Wikipedia practice and therefore does not have consensus).

PSTS is basically concerned with describing the sources, their strengths, and their weaknesses. NOR is basically concerned with which claims in the sources can support which claims in the articles, in particular, avoiding improper synthesis. This means using factual claims to support the same factual claims and using interpretations to support the same interpretation. So NOR has to classify statements as factual or interpretive, and probably also note issues which make for unreliable statements in reliable sources, while PSTS has to classify sources as Primary, Secondary, or Tertiary, with some special cases. (P.S. Count me in groups 1 & 2. I initially supported keeping PSTS in a "self-contained" NOR policy, but I think the new proposals cover the NOR issues and mean we can move PSTS elsewhere; I now think PSTS would do much better on its own or with other source-classification policies). Jacob Haller 15:53, 23 September 2007 (UTC)

I agree in principle, and think we can avoid classification altogether (see draft above). Rather than classify the sources broadly, we can have special provisions for special cases, like "ancient sources", which should be called what they are. Dhaluza 16:06, 23 September 2007 (UTC)

Reply to jossi: The requirements for demonstrating consensus are not the same as the requirements for demonstrating lack of consensus. I would argue that the arguments of 5 or 10 editors does demonstrate lack of consensus. What do you propose as an alternative? i.e. is there an existing or previously existing version of the policy that has broad support? --Coppertwig 16:26, 23 September 2007 (UTC)

The arguments of 5 or 10 editors only demonstrates that they disagree with long-standing material, The burden is on them to gather sufficient consensus to change established policy. This should be obvious. ≈ jossi ≈ (talk) 17:36, 23 September 2007 (UTC)

If, and this is a big if, there was a way through transclusion to accomplish change, while still keeping what some people deem as "imperative to be included", would alleviate "group 3's" concerns? For example, with transclusion, it would be possible for the entire PSTS section to be own it's own page (call it maybe something like a sub-policy, or maybe just policy on it's own). Then, using transclusion, that 'page' can be brought in to this policy as well, to still be part of this policy (even though I'm in the camp that doesn't think it needs to be). Any further discussion or changes, etc. can be referred to that page, while at least this talk page can maybe become devoted to solely NOR issues, without being 'cluttered' with discussions on types of sourcing, or other "source" related discussions. Just looking at this page and the last two archive pages clearly show that this would help clean up the talk page here. By transcluding, the disputed text is still kept within the policy, just that it's stored differently, and therefore changes or discussions can be handled differently as well.

Just an idea I'm tossing out for the moment for possible consideration. This could be used as one way to at least begin trying to address what 2 different groups see as a problem, even if the thrird group doesn't think any exists. Their needs or concerns are still addressed by keeping the text they think is so important. By transclusion, the text is brought in and becomes part of the "article", or "policy" in this case. wbfergus ^Talk 16:42, 23 September 2007 (UTC)

<undent> The proposal to make a complete package... hammering out a proposed PSTS policy/guideline page to go along with the proposed replacement section for NOR, looks welcome to me. What's needed with it is a FAQ showing the evidence for the concerns, for example the contexts in which definitions vary, and diffs of example of disputes over interpretation. We also need to consider carefully the policy implications of PSTS and of the proposal, both in this policy and in relation to other policies. There could be good use of a page covering both this and what exactly is intended by "third party", possibly with text brought in through transclusion as above. .. dave souza, talk 17:38, 23 September 2007 (UTC)

There is nothing to stop good intentioned editors to start a proposed page to cover these issues, designed with a possible future transclusion in mind. ≈ jossi ≈ (talk) 17:40, 23 September 2007 (UTC)

On the page Talk:Ebionites, the following statement was made, and I quote: "Citing two authors with divergent views in some areas is not OR nor synthesis. What matters is whether they are making the same claim in the specific area which they are cited about -- which in this instance they are". Is that statement in accord with wikipedia policy as per this page? John Carter 16:54, 23 September 2007 (UTC)

I'm not going to comment on the actual dispute, which seems quite heated. My observation is that if there are two sources saying the same thing, then quoting that thing from two sources is not synthesis, instead you have a corroboration of different sources. If two sources, assuming they would be thought of as reliable sources, that differ on their interpretation elsewhere concur on some point, then that would be pretty good evidence of agreed facts.

The devil is in the detail though, and the synthesis can often be not in the citation of an agreed fact, but what use is then made of it. Another question to answer is, is the citation really in context and being used as the author intended, or is it some passing aside being extrapolated into support for some novel argument? I am always deeply suspicious of debates that depend on policy over editorial judgement, and I would be looking at whether it was an issue of the neutral point of view policy. Spenny 17:07, 23 September 2007 (UTC)

It unfortunately appears to be the case that WP:PSTS has recently been manhandled by too many various, unsynchronized edits in the past few months, and is in need of work. As I stated farther above, "Primary and secondary sources" has been a standard part of WP:NOR for some three-and-a-half years, with its roots in direct instruction from WP's founder. A couple years ago it became a discrete section of WP:NOR. In October 2006 it was changed to "Primary, secondary and tertiary sources", which solved many conceptual problems and gave an extra layer to the concept of "secondary sources" in the implementation of WP:NOR around the wiki. Nearly a year ago WP:PSTS read as can be seen in the version in late October 2006. Notice also how it still read pretty much the same in April 2007. In the interim a some "entropy", so to speak, has occurred. Note how it began to be worked over gradually, bit by bit, e.g., by 23 June 2007, with the example of the "Security Resolution". By the beginning of August 2007 it looked like this. Presently it looks like this. This section apparently only needs to be cleaned up and brought back to its earlier, more stable roots, not thrown out. ... Kenosis 17:15, 23 September 2007 (UTC)

Yes. ≈ jossi ≈ (talk)

See Dave's proposal above, which I would support, in which well intentioned editors could start a proposal page to cover PSTS, with the intention to be proposed as a transclusion for this and maybe other pages. ≈ jossi ≈ (talk) 17:42, 23 September 2007 (UTC)

No objection to a transclusion approach, so long as the participants there don't get too far off into manufacturing unwieldy categorizations. The PSTS distinction is supposed to be a way of conceptualizing the flow of information directly integral with the concept of NOR. In other words, the objective is to avoid creating primary sources or new syntheses in WP. The tertiary distinction is just shorthand for, essentially, multiple layers of synthesis in published material. ... Kenosis 18:23, 23 September 2007 (UTC)

By far, the worst statement at present is the bolded statement: "Wikipedia articles should rely on reliable, published primary and secondary sources." This statement was originally added on 28 June 2007 in the midst of a brief flurry of edit warring by several users. It originally read: "Wikipedia articles should rely on reliable, published secondary sources." It was arguably an oversimplification of "secondary sources" from the very beginning of that bolded policy statement in late June, The statement, with the addition of "primary and..." to the sentence, has since become confusing and essentially meaningless. It's no wonder a number of people are more confused by PSTS than need be the case here. IMO, the statement should never have been added to begin with, or at an absolute minimum it should be set into better context. ... Kenosis 17:57, 23 September 2007 (UTC)

That edit must be reverted. It was widely challenged and was added again against consensus. ≈ jossi ≈ (talk) 18:00, 23 September 2007 (UTC)

Going back to an older version would go a long way towards clearing up my concerns... somewhere along the lines we lost a key statement (that: "Original research that creates primary sources is not allowed.")... this placed the discussion of sources in the context of NOR. It made it clear that the point was to not add things that made Wikipedia a primary source. While I still think the idea of having all this in a seperate policy/guideline has merit, I could support Kenosis's idea as an acceptable solution while a proposed version of such a page is created an goes through the process of determining consensus. Blueboar 18:11, 23 September 2007 (UTC)

Going back through the Archives, it does appear that there has much discussion over this section, going at least as far back as January 2005 [[3]]. What I find unusual so far, since I haven't yet gone through all of the archives, is that in this case, SlimVirgin is making the same points about primary sources that many editors still feel important about now, yet SlimVirgin is also the editor that made the numerous edits back in Oct 2006 with any discussion. I will be interested to see what arguments were finally made to persuade SlimVirgin to change their position. They may be viable to bring back up. wbfergus ^Talk 18:20, 23 September 2007 (UTC)

It's definitely true that we'll find little talk-page material on SlimVirgin's 23 October 2006 change from PSS to PSTS. What is clear is that it was sufficiently effective at solving certain conceptual problems, e.g., specialized encyclopedias, general encyclopedias and other summaries of what had been grouped as "secondary sources", such that there was a more-or-less collective "aha!" at the time. Hence, no talk-page discussion was needed there. Also, the farther back we go, the less talk-page discussion we'll find in general right across the wiki. I agree the amount of talk last October was not in keeping with the ordinary expectation. Moreover, I am not an opponent of going back to PSS instead of PSTS, so long as compendia, summaries and various derivatives of other secondary source material are effectively dealt with. The main reason I see for not going all the way back to PSS is the issue of encyclopedic matter built on secondary sources. If someone comes up with an improved way of comprehensively dealing with the issue of, shall we say, "secondary sources built on other secondary sources", I figure others will tend to recognize its potential value and discuss it accordingly. If Dave Souza gets his way, there may be a separate page to work on that one issue. But it must be interactive with WP:NOR, because WP not being a primary source is integral to NOR.. ... Kenosis 18:39, 23 September 2007 (UTC)

The addition of "primary" to that clause does two things: it contradicts the guidance on using primary sources; and creates confusion. What the editor who added primary to that clause thought he/she was doing is unclear, but it needs to be edited as Kenosis has noted. 18:22, 23 September 2007 (UTC)

As the link I just provided above shows, this issue has been in contention for quite a while now. wbfergus ^Talk 18:25, 23 September 2007 (UTC)

But what guidance is there, with consensus, that says that primary sources being used is a problem in and of itself, rather than just something that needs more/different caution than secondary sources? SamBC(talk) 18:37, 23 September 2007 (UTC)

Yes, of course. To go back to the early roots of this concept, prior to when the various practical experiences with it gave rise to the manifold PSTS general distinction, see the following: Jimmy Wales 3 December 2004 and Jimmy Wales 6 December 2004. This may help to focus on the primary objective of PSTS. ... Kenosis 18:59, 23 September 2007 (UTC)

That indicates certain problematic uses of primary sources, but does not indicate any general problem with every using primary sources generally. Plus, while Jimbo's opinions tend to lead consensus, Jimbo saying a few things doesn't automatically indicate consensus guidance. Sure, don't use primary data to create novel theories. That's valid and no-one's going to argue with it. If you remove everything after "data" then it becomes rather more general and contentious. SamBC(talk) 19:11, 23 September 2007 (UTC)

And, in between then and now, there is a longstanding consensus on the basic aspects of its development in the intervening three years. At this stage, the onus is upon those proposing any signifcant changes. ... Kenosis 20:06, 23 September 2007 (UTC)

Whether the consensus was londgstanding, or even was, is only of historical interest now that it has been widely and thoughfully questioned. This latest turn in the discussion seems to me to be like trying to put the genie back in the bottle. The two bolded statements at the top of this section were indeed the root cause of the edit war. My POV is that statement depreciating primary sources did not represent consensus, and that inserting it over the objections raised opened the whole can of worms on PSTS. My attempt at compromise language was immediately reverted, for example. Frankly, if editors would have agreed to just leave the bolded statement out altogether, the edit war probably would have died down, and we would not be at the point we are. But since it was reinserted every time the protection expired, and that lead to re-protection, the talk page discussion has questioned the whole concept of the relevance of PSTS to NOR, and I have not seen a cogent stand alone argument for it other than it's been in there for X amount of time and we really really really need it. That's a religious argument in my book, and carries little weight. I could be swayed by a cogent argument, but I have not seen any yet. I have made a proposal on how to reword PSTS to remove source typing above, and I think it is actually forms a stronger policy because it is more easily understood. Explain why that explanation is inferior to primary/secondary. Dhaluza 20:24, 23 September 2007 (UTC)

How about this:

Most succinctly,

• primary sources are sources of facts
• secondary sources are sources for distinct views of facts
• tertiary sources are summaries of, or generalizations based on, diverse views of facts

More specifically:

• Primary sources record data that are to be interpreted or explained, or data that are used to interpret or explain other data. Examples of primary sources include archeological artifacts; photographs; historical documents such as diaries, census results, video or transcripts of surveillance, public hearings, trials, or interviews; tabulated results of surveys or questionnaires; written or recorded notes of laboratory and field experiments or observations; and artistic and fictional works such as poems, scripts, screenplays, novels, motion pictures, videos, and television programs. Note: often times secondary sources present data and may thus also function as primary sources.

Our policy: Original research that produces a primary source is not allowed at Wikipedia. Moreover, a Wikipedia article or section of an article can employ primary sources only if the source is used (1) only to make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge, and (2) never to make original analytic, synthetic, interpretive, explanatory, or evaluative claims. Contributors drawing on primary sources should be careful to comply with both conditions.

What do people think? I'd like to add something to make it clear that if we are citing a secondary source and that source includes raw data, we can include the raw data as long as it is restricted to the context in which the authors of the secondary source (e.g. journal article) use it i.e. in the context of their own analysis/argument ... but I haven't figured out a concise way to say it... Slrubenstein | Talk 19:26, 23 September 2007 (UTC)

1. I understand the addition of primary sources to the bolded statement as a response to the non-standard-practice deprecation of primary sources implied by the bolded and italicized statements.
2. I suspect that the requirements of explaining PSTS and the requirements of using PSTS to explain NOR conflicted with each other, leading to more entropy; using non-PSTS explanations of NOR would avoid this. Jacob Haller 20:16, 23 September 2007 (UTC)

I think we should avoid using the words primary and secondary in explaining NOR. The arguments against source typing are compelling, and the need for this characterization has not been shown. The problem with NOR is the use of the sources, not their nature. Dhaluza 20:28, 23 September 2007 (UTC)

I still do not support any changes to the existing policy as it is, but if a compelling case were made that changes are necessary (and one hasn't yet) SLR's are reasonable and accurate enough. FeloniousMonk 20:40, 23 September 2007 (UTC)

I would agree with that as well. ≈ jossi ≈ (talk) 21:43, 23 September 2007 (UTC)

What's currently missing from Slrubenstein's proposal for shortening PSTS is that WP is not permitted to be a publisher of original syntheses, and as such is not a publisher of original secondary source material either. The primary/secondary/tertiary distinction helps clarify this. BTW, I liked the basic idea that Blueboar proposed of "Sticking to the sources". Any ideas about how a very brief articulation of this might be woven in without interrupting the continuity of basic policy across the wiki? ... Kenosis 20:46, 23 September 2007 (UTC)

"What's currently missing from Slrubenstein's proposal for shortening PSTS is that WP is not permitted to be a publisher of original syntheses," - how can you say this? i address this explicitly! Slrubenstein | Talk 22:04, 23 September 2007 (UTC)

The way this flows appears not to express the concept at first reading. It says: "Our policy: Original research that produces a primary source is not allowed at Wikipedia" That appears to neglect original synthesis, which is also original research. Or am I missing something in this proposal? ... Kenosis 23:26, 23 September 2007 (UTC) ... OK, I see it now, in #2. Fine, no problem. it draws on existing long-used language and doesn't appear likely to confound users who have relied on the older language for any length of time. It's not a major change, but a brief clarification, if I'm looking at it correctly. ... Kenosis 23:32, 23 September 2007 (UTC)

Regarding wikipedia not being a primary or secondary source... it would make sense, to my mind, to pretty much completely rewrite the PSTS section so that its main thrust is defining all three and stating that wikipedia is meant to be a tertiary source. The corrolaries of this belong in other sections of the policy, to my mind. This would not implicitly or explicitly prefer primary or secondary sources over the other, handily. I would personally favour, though, a seperate page (of policy standing) defining PSTS, and just referring to that and saying "wikipedia is supposed to be a tertiary source". However, the definitions can stay in NOR while that new page is developed, and be replaced with a wikilink reference when the new page is adopted. The new page would also be a good place to keep all the definitions of different types of source (including third-party) needed by policies and guidelines, and refer to them there. It would need to be of policy standing itself because it would be a definition officially recognised and adopted. A new term might be worth using, other than policy/guideline, such as "policy support", but that would most likely over-complicate things. Anyway, consider that a suggestion. If people don't seem to notice I've made it in the middle of all this discussion, I'll start a new section for it... SamBC(talk) 21:46, 23 September 2007 (UTC)

When an article or part of an article draws on primary source material, it is reasonably regarded as a secondary source. When it draws on secondary material, it is reasonably regarded as either a tertiary source, or a derivative secondary source, depending on whether you're using the old PSS or the more recent PSTS breakdown. The prohibition is against being a primary source, and also against synthesizing primary source material in a way that creates an original secondary source. ... Kenosis 22:00, 23 September 2007 (UTC)

Making use of a primary source doesn't render the result a secondary source as an automatic consequence - according to Tertiary source, "a tertiary source is a selection and compilation of primary and secondary sources" (bolding in original). SamBC(talk) 22:05, 23 September 2007 (UTC)

Sure, SamBC, Note that you just quoted a Wikipedia article on Tertiary source. If you know what you're doing here, I would prefer to not continue feeding back on this sort of junk argument on a policy discussion page. If you don't know what you're doing here, then please learn better. I shouldn't need to explain this sort of thing on a policy discussion page. ... Kenosis 23:00, 23 September 2007 (UTC)

I'm not sure what the problem with quoting that article is, given that the article is also linked from the relevant bit of this policy page (presumably to give further information). The policy itself also says that tertiary sources sum up primary sources as well as secondary. This university-hosted glossary also says the same. You might also like to keep your comments more

WP:CIVIL. All this is relevant assuming that your comment that I quoted in response to was actually a response to my suggestion to reword the section to focus on wikipedia being a tertiary source. Otherwise, we've somehow started talking at cross-purposes. SamBC(talk

) 23:37, 23 September 2007 (UTC)

That comment was definately not called for and seriously broached both

personal attacks, not to mention just downright common courtesy. Sam's lack of response to that uncalled for comment shows maturity and civility that is missing in some posts on here, that seem (on the surface at least) to be no more than obstructionist tactics with strawman arguments. Claiming that 'something' wasn't widely challenged for a period of time doesn't neccessarily mean that there was widespread concensus for it. If that was truly the case, the wouldn't have been an Emancipation Proclamation or a movement towards Abolitionism

.

For the record I would like to point out that it has been widely acknowledged that PSTS is a breif synopsis of the definitions of the individual P,S, and T articles, even though those articles may have come into existence after these concepts were argued on this page (I haven't checked). The NOR policy page does not get into the full definition of those, otherwise it would be completely pointless to have a full 'policy' definition of them and still link to the articles with complete definitions, explanations and examples. wbfergus ^Talk 23:34, 23 September 2007 (UTC)

And I could very easily argue that continuing to bring up specious or irrelevant points is uncivil, in that doing so wastes our time. Kenosis' assessment of the situation was correct. &#0149;Jim62sch&#0149; 11:47, 24 September 2007 (UTC)

And what would the "specious or irrelevant points"? That almost all of the discussion at least over the last month (actually a little longer now, that I know of), has revolved almost exclusivley around the current "Sources" section, primarily PSTS? Almost all of this page, and almost all of the last two archive pages show that this is where the bulk of conversation has been.

Or did you mean instead that unlike a previous edit or two (or many more) were done in clear violation of existing Wikipedia policy, as can be readily seen by the dates and versions for those dates, and so that we don't fall into the same sort of quagmire a year from now of somebody asking "Why wasn't policy followed for these edits that are now so contentious", we can instead show that policy is being followed. There is a very healthy and long debate going on with how to address the issue of moving extraneous material that has nothing to do with "No Original Research" someplace else, so that this policy can concentrate solely on NOR issues. It is ludicrous that a policy devoted to "No Original Research" is instead expending most of it's time and energy deabting the differences between primary, secondary and tertiary sources. Nobody (that I can tell) is proposing weakening any of the existing policies, but simply moving this extremely problematic section somewhere else so that any further discussion about the various types of sources or "this can be a primary in this case, but it is also a secondary in this case", etc. type discussion do not clutter up something that should be devoted to NOR issues. Why should NOR be concerned with whether a source is primary or secondary or tertiary? Why can't editors simply look at the problemmatic text from an article and ask "Is there a

reliable

source for this statement or conclusion or whatever the statement is about"? The type of source, while related, is not a core part of this policy, though as has been previously stated, it's been part of the policy since the 23rd edit. As my analogy above stated, just because something has "been that way for a longtime", doesn't make it correct or the best way.

Or maybe were you referring to my comment that some posts above can be seen as obstructionist? When asked what they think, some people just say "Object" (or a variation of it), and when asked how it could improved to alleviate their concerns, we (those who want to clarify and simplify the policy for easier understanding by all Wikipedians, both old and new) either have our questions met with silence, answered with a variation of "it's been long-standing policy....", or other useless suggestions or comments. There have been extremely few people (Slrubenstein and Jossi that I remember right off-hand) from the group that has voiced opposition that actually appear to be open-minded about the issue and have actually bothered to offer constructive criticism instead of obstructive criticism. wbfergus ^Talk 12:58, 24 September 2007 (UTC)

I believe SamBC, upon meditating on it for awhile, will understand what he did (not for the first time, incidentally) and will also understand what I meant just above. As to the hasty exaggerations about "violations" of

WP:NPA, I stand by what I said. I'll drop SamBC a note via email in a couple of days and I presume we'll briefly discuss it in private. If we need to publicly analyze it in greater detail, I'm prepared to spend some further time doing it on a separate page. Other than some additional indicator that it's necessary here, I consider this ancillary issue closed as far as the purposes of this talk page are concerned. ... Kenosis

13:13, 24 September 2007 (UTC)

Let's remember the point of all this explaining of source types... which is to say that Wikipedia should not be 1) a primary source of a fact or idea... or 2) an original secondary source in analyzing facts or ideas. IE facts and ideas should not originate on wikipedia. All else is fluff to help explain these two concepts. Blueboar 22:37, 23 September 2007 (UTC)

I see nothing wrong with "Wikipedia articles should rely on reliable, published primary and secondary sources." UN Security Council resolutions are published primary sources. What is not wanted is unpublished primary sources. What is not clear to me is whether manuscripts only held in one archive are acceptable. I asked this question either on this talk page or the talk page of WP:V when I came across it on the Chindits page and the people who bothered to answer seemed to think it was OK if they were accessible by the public. --

Philip Baird Shearer

19:48, 24 September 2007 (UTC)

That sentence was at the heart of the edit war that started this whole mess. Some editors want it to only include secondary sources thereby depreciating primary sources, while others disagree, and some think it should also include tertiary sources. On top of this, editors can't agree on what the precise definitions of PSTS are, or how to apply them to specific examples. So there is clearly no consensus support for this statement in any of its forms, and removing it is the best alternative. Dhaluza 10:25, 25 September 2007 (UTC)

The problem with the sentence is it misses the point of NOPR which is not simply that people must use verifiable sources, which is covered in V and CS - the heart of NOR is how people use sources, and misuse them. The main ways people misuses sources is

• by taking primary material (whether from a purely primary source, like a book of French poetry, or a secondary source, like a journal article analyzing a French poem that provides the text of the original poem; or like a biological assay, or data from a biological assuay included in a peer-reviewed journal article using that data) out of context or
• combining material from primary sources to forward their own views

and the most obvious form of NOR was to do either of the above in order to argue against a view in a reliable, verifiable, secondary sources. Now, it is clear to me that to explain this, we need to make some kind of distinction about sources, and I grant this is not easy. But the point of it all has to do with the ways people use sources. It is not enough just to say, "use sources." NOR is about the fact that there are wrong ways (for Wikipedia) to use them. Slrubenstein | Talk 10:40, 25 September 2007 (UTC)

We had concensus for policy changes?

In case people don't have a "Watchlist" set for the policy page itself, it automatically was unprotected this morning by the 'bot, and two edits have taken place. Both covered an area that's been under discussion, but I don't remember seeing anything that asks for if people agreed or not.

• One edit removes "An eyewitness account of a traffic accident is a primary source. United Nations Security Council resolutions are primary sources.".

Agree wbfergus ^Talk 13:13, 24 September 2007 (UTC)

Don't disagree... but wonder why these were removed? Aren't they primary sources? Blueboar 13:18, 24 September 2007 (UTC)

I removed the UN Security Council Resolution and Encyclopedia Britannica examples, because they are not broadly enough illustrative of primary and tertiary sources. they made things confusing by omission of other representative examples. I recommend instead coming up with an agreed list of examples of P, S, and T sources and listing them with bullets, or alternately to leave the unnecessary examples out completely. ... Kenosis 13:28, 24 September 2007 (UTC)

Maybe in an attempt to begin removing the definitions of "types of sources" so emphasis is instead on puerly NOR isssues? wbfergus ^Talk 13:24, 24 September 2007 (UTC)

• Another edit removes Wikipedia articles should rely on reliable, published primary and secondary sources.

Agree, though I think the same could have been accomplished with just removing "primary and secondary".wbfergus ^Talk 13:13, 24 September 2007 (UTC)

Agree - no need to repeat what is stated earlier in the policy. Blueboar 13:18, 24 September 2007 (UTC)

• Agreement on adding Slrubensteins "Origins of NOR"?

The way I read the "unofficial" vote at the top of the page, it appears that there is concensus for that, so I will add it. If it turns out there is not concensus, then go ahead and revert. wbfergus ^Talk 13:13, 24 September 2007 (UTC)

Yes, it appeared plain that the insertion of this passage earlier this year and the later modification of it was terribly confusing. So was calling WP a tertiary source. In some cases WP is a "secondary" source and in some cases a "tertiary" source, but never a primary source, at least not according to the NOR policy. ... Kenosis 13:28, 24 September 2007 (UTC)

As of yesterday, Blueboar's proposal read:

Citing Appropriate Source Materials (alternative suggested section title: Sticking to the Sources)

Within Wikipedia articles we will find statements of fact and statements of interpretation, analysis or conclusion. It is important to cite appropriate sources to back those different types of statements. Statements of fact should be cited to reliable sources that contain that fact. Statements of fact should also match the context of the source for that fact. Statements of interpretation, analysis or conclusion should be cited to reliable sources that contain the same interpretation, analysis or conclusion, often called secondary sources (see

). Collections of facts which tend to lead the reader to a certain interpretation should be supported not only by a reliable source for each fact, but also by a reliable source that contains the same interpretation.

Editors cannot include their own interpretations of previously-published facts, unless that interpretation is either 1) an obvious and non-controversial consequence of the facts or 2) can be attributed to a reliable source. Nor can editors expand on an author's interpretations of fact, unless that expanded interpretation is also found in a reliable source.

When there are a number of reliable sources that interpret a particular piece of material, we need to be especially careful not to insert or imply our own interpretation of the original material.

For example, that Franklin D. Roosevelt was born in Hyde Park, New York and that he became president of the United States are both verifiable facts that may be cited in a Wikipedia article. The dates of each event are easily verified, and it would be perfectly admissible to say that he "was born in Hyde Park, NY, and later became president of the United States" without finding a source that specifically uses the word "later", or otherwise explicitly compares the dates. However, any statement about the effect his birth place had on him or on his career would require separate citation, since such a statement would not an obvious or non-controversial consequence of these easily verified facts.

Comments & concerns

Support I think it summarizes NOR as cleanly as any of the alternatives. I hope that we can work out any remaining issues and add this to the policy by, say, the end of the year, regardless of how we handle PSTS. Jacob Haller 21:19, 23 September 2007 (UTC)

I'm not sure that we should pursue my proposed new section without addressing the PSTS section at the same time... unless you are suggesting adding it on top of PSTS(?) I don't object, if that is the consensus, but it wasn't written with that intent. Blueboar 21:29, 23 September 2007 (UTC)

I can see value in adding it seperately at this point, and continuing the discussion of refactoring or relocating PSTS. I would prefer both refactoring and relocating, personally. SamBC(talk) 21:38, 23 September 2007 (UTC)

OK. Blueboar 21:43, 23 September 2007 (UTC)

Note: I have added a sentence about not taking facts out of context... an easy way to introduce OR with a fact statement. Blueboar 21:51, 23 September 2007 (UTC)

• Where is this proposed to be placed? In lieu of something? As an addition to some existing section? As a separate section? ... Kenosis 22:05, 23 September 2007 (UTC)
  • I would suggest immediately before or after the PSTS section, with the understanding that the refactoring/rewriting/moving of PSTS is still under discussion. SamBC(talk) 22:08, 23 September 2007 (UTC)

I support this as a plain language explanation of NOR, but think the reference to

WP:PSTS is unnecessary. Dhaluza

22:49, 23 September 2007 (UTC)

Dhaluza, could you accept that, even if it's unnecessary, it does no (substantial) harm, and therefore not oppose its inclusion? SamBC(talk) 23:40, 23 September 2007 (UTC)

Right now, I would tend to support it. Maybe after I eat and sleep I may change my mind, but for now it seems a completely logical extension. It by far seems to make the point about NOR better than PSTS does, but that is still a separate issue. wbfergus ^Talk 23:55, 23 September 2007 (UTC)

Support - Let's add this and Srubenstein's context/history to the policy without deleting anything. Can we all agree on that? Let's get some kind of improvement into the policy. WAS 4.250 07:39, 24 September 2007 (UTC)

Support with one modification. I agree with

18:20, 24 September 2007 (UTC)

Cogden... I also support removing the paranthetical... but not for the reason you state. Quites simply, you have unique definitions of the terms Primary and Secondary source. If we avoid the terms, then this does not matter... but they are unique. Blueboar 18:31, 24 September 2007 (UTC)

It's not unique. I guess it's time to bring back my numerous citations to sources that define primary and secondary sources that way. That's what happens when you get so much activity on a page. Previous arguments and citations get swept into the archives before people see and digest them. COGDEN 22:37, 24 September 2007 (UTC)

Oppose this: "Editors cannot include their own interpretations of previously-published facts, unless that interpretation is either 1) an obvious and non-controversial consequence of the facts" which is a departure from the policy and would require far more support to fly than the number of editors currently involved in this discussion. Slrubenstein | Talk 18:25, 24 September 2007 (UTC)

I don't see it as a departure... Take a look at the example... isn't saying something like "FDR was born in Hyde Park and later became President" allowable? Are you saying that such a statement can not be made without a source? Blueboar 18:36, 24 September 2007 (UTC)

It doesn't seem to be a departure from the meaning, spirit, or intent of the policy. However, I can understand why there'd be concern about it potentially causing problems. I just don't think that people will get far trying to stretch what "obvious and non-controversial consequence" means. SamBC(talk) 18:49, 24 September 2007 (UTC)

This is nothing new. It's part of

WP:V. If something is obvious or non-controversial, you don't even need a citation to a reliable source! All the above formulation says is that editors cannot include their own interpretations unless they are verifiable. That Wikipedia practice and consensus has been well established for years. COGDEN

18:50, 24 September 2007 (UTC)

Eh, editors cannot include their own interpretations. They can only make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge. Editors can summarise sources of fact, taking care not to introduce an interpretation and change the meaning by a selective summary. Or, as the term goes,

quote mining. Any interpretation must be an accurate summary of the interpretation of a third party in a secondary source (in relation to the subject matter of the article or section). Sorry if you find that concept difficult. .. dave souza, talk

21:30, 24 September 2007 (UTC)

What I'm saying is that if the editor cites a reliable source that agrees with an editor's interpretation, the editor has not committed original research. That is, her idea isn't new. They can't cite themselves, in any case; they have to cite a reliable source. But just because the editor came up with the idea doesn't make it bad. What's bad is that the editor was the first to come up with the idea, and/or the editor doesn't cite a verifiable source. COGDEN 22:34, 24 September 2007 (UTC)

Agree with COgden, with two minor technical clarifications. A WP editor can cite their own work if it was published in RS. Also they don't need to be the first to come up with it, they need to use WP to publish it before it appears in RS to be doing OR. Dhaluza 09:44, 25 September 2007 (UTC)

How about this as an alternative: "Editors cannot include their own interpretations of previously-published facts. Editors may contextualize facts in an encyclopedic way that is an obvious and non-controversial consequence of the facts" By refocusing on context rather than interpretation, we stick to what is necessary for an encyclopedia, rather than going down the slippery slope of what is not. Dhaluza 09:50, 25 September 2007 (UTC)

Alternative draft

Using Appropriate Source Materials Appropriately

Wikipedia articles contain statements of fact, and statements of interpretation, analysis or conclusion. It is important to cite appropriate sources in each case.

• Statements of fact should cite reliable published sources that state the same fact in an equivalent context.

• Statements of interpretation, analysis or conclusion should cite reliable published sources that contain the same interpretation, analysis or conclusion (also called secondary sources; see:
  WP:PSTS
  ).

• Collections of facts presented in a given context should be supported not only by a reliable source for each fact, but also by reliable sources for the context.

Editors cannot include their own interpretations of previously-published facts, unless that interpretation is either 1) an obvious and non-controversial consequence of the facts or 2) can be attributed to a reliable source. Editors can only expand on a source's interpretations of fact by citing additional reliable sources in an appropriate context.

When there are multiple reliable sources that interpret a particular concept differently, editors need to be especially careful to present all mainstream and significant minority views, and not develop a novel interpretation or reinterpretation of the original material. When the collection of facts support different interpretations, editors must present the relevant verifiable facts, and not selectively present facts that support one interpretation over another.

For example, that Franklin D. Roosevelt was born in Hyde Park, New York and that he became president of the United States are both verifiable facts that may be cited in a Wikipedia article. The dates of each event are easily verified, so it is reasonable to say that he "was born in Hyde Park, NY, and later became president of the United States" without citing a source that specifically uses the word "later", or similarly characterizes the sequence. However, any statement about the effect his birth place had on him, or on his career, would require a citation, since that would not be an obvious and non-controversial consequence of these facts. Both of these facts are important biographical information that should be included in his biographical article--an editor who thought Roosevelt was a bad president should not simply ignore that important part of his life.

Comments & concerns

Apologies in advance if this seem shocking. I broke out the ax and chopped down excess wordiness. In the process I also changed a few things. The title is new. I broke the paragraph making two different points into bullets. I refocused the third statement on context rather than the slippery slope of tending to lead to... because I could see this being used to reinvigorate controversy over lists for example--instead, I added a sentence about cherry picking facts, which should address the same problem with less collateral damage. I expanded the example to tie it to the last paragraph as well. Feel free to compare and contrast this with the original proposal (I have thick skin). Dhaluza 10:38, 24 September 2007 (UTC)

Likewise editors cannot expand on a source's interpretations of fact, except by citing additional reliable sources. I can see your troubling over that too. This has the potential to suggest synthesis is ok (there is a fact, there are some other facts, look I can join them together), and if additional sources are cited, then the interpretation depends on those so is not an issue. My instinct was simply to delete the second half of the sentence, but perhaps some wording is required to hammer home the point. It is also ambiguous for not saying what these additional citations are required for. Spenny 10:48, 24 September 2007 (UTC)

Yes, the original sentence is redundant, so I tried to make it relevant, but I see your point too. Oh well, it's still a draft on a talk page, so we can kick it around the block again.... Dhaluza 10:52, 24 September 2007 (UTC)

Sorry, I should have let it set :) I think the current wording is better, but still has that problem of being "too helpful". I could paraphrase it to say "You can only extrapolate from the facts if you can cite that extrapolation." That still seems redundant. Thinks, how about "Editors may never make their own extrapolations from cited information."? Spenny 10:59, 24 September 2007 (UTC)

No good, it's just repetition really isn't it? Spenny 11:00, 24 September 2007 (UTC)

OK, I see the subtlety, how about: Editors cannot include their own interpretations of previously-published facts, unless that interpretation is either 1) an obvious and non-controversial consequence of the facts or 2) can be attributed to a reliable source. Any interpretation must be completely contained within the cited source. Spenny 11:15, 24 September 2007 (UTC)

I like this more than Blueboar's version, though I still have doubts about the Roosevelt example. Slrubenstein | Talk 18:25, 24 September 2007 (UTC)

I agree much of this version, though I think Blueboar's version is more likely to achieve consensus. I have the same issue as with the other proposal, with the parenthetical "(also called secondary sources; see: WP:PSTS)". An analysis, interpretation, or conclusion is not called a "secondary source" (not by academics, anyway). If the analysis is original, it's actually the primary source of that analysis. We don't need this parenthetical, and it's potentially confusing.

I also have serious qualms about requiring "reliable sources for the context" in the OR article. Obviously, we want editors to provide context for quotations and facts, but that's mainly an

18:46, 24 September 2007 (UTC)

It becomes an NOR issue when someone takes something out of context to forward their own argument. Put another way, taking things out of context is one of the most common ways people violate NOR. Slrubenstein | Talk 18:55, 24 September 2007 (UTC)

My point is that simply citing verifiable facts is never original research. If the fact is verifiable in the first place, then merely omitting some other facts or context is not OR, just bias. If you disagree and think it's possible to commit OR merely by omitting context on a verifiable fact, I'd like to see an example. COGDEN 20:36, 24 September 2007 (UTC)

See

quote mining. ... dave souza, talk

21:40, 24 September 2007 (UTC)

In quote mining, the quotes themselves are not OR. The OR (actually NPOV in this case: I still don't see any examples of OR) comes from the commentary, such as when you say, "Quote X says Y, but as you can see from Quote Z, the opposite is true". Quote X and Quote Z, by themselves, are just quotes. They are not new research. COGDEN 22:27, 24 September 2007 (UTC)

COGDEN, original analysis and/or synthesis does not mean it is a primary source in the humanities. On the contrary, published articles by experts in humanities are expected to contain novel points and yet considered secondary sources. Similarly, historical texts are generally considered primary sources in humanities, regardless of whether they would fit a "technical" definition of secondary or tertiary. Additionally, some historiographers and philosophers would even be "extreme" enough to say that any material you draw upon for your study is a primary source. There is no single universal and coherent definition for the distinction between source types. Vassyana 20:02, 24 September 2007 (UTC)

They are considered secondary sources as to the previously-published material they reference. (As to the novel insights based on those prior sources, they are primary sources.) It's very well established in the field of historiography, where the subject matter is historical ideas, that the primary sources include the historian's novel analyses of other primary sources. The historian to come up with a novel historical idea is the primary source as to that idea. See, this issue is so advanced and difficult, even for very intelligent people like Wikipedia editors, that its hard to justify basing a Wikipedia policy on the fine points of distinction between primary and secondary sources. That's PhD stuff, not really a distinction upon which Wikipedia should hinge. COGDEN 20:36, 24 September 2007 (UTC)

To be entirely blunt, you don't know what you're talking about. You are confusing fields, in terms of the typing of sources. First, historiography is not at all synonymous with with the disciplines of history, anthropology or archeology, and it additionally has a proprietary set of terms, like most specialist fields. Second, the primary/secondary distinction is only applied to "symbolic" sources. Primary sources in historiography are exclusively texts that originate in the time period being studied, providing a real and direct link to the historical period. Secondary sources are later sources that draw upon these earlier reference. In historiography, an ancient secondary source is still a secondary source. A modern secondary source is still a secondary source irrespective of novel content or theories. In that field, a modern source is only a primary source if it comments upon modern circumstances and is a symbolic source. It has nothing at all to do with making novel claims, at least in historiography. Vassyana 22:42, 24 September 2007 (UTC)

<undent> Again, editors cannot include their own interpretations of previously-published facts, but can only make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge. .. dave souza, talk 21:40, 24 September 2007 (UTC)

I think the excange above is actually a very good example of why using primary and secondary makes it more difficult to explain and debate NOR. It started out on a valid exchange on how statement of facts can or cannot be OR, and went off into a side discussion of primary vs. secondary instead. These distinctions may be of academic interest, but I think they have proven to be of little practical value in writing an encyclopedia that anyone can edit. We should stop using them in this discussion, and banish them as OR for not being "easily verifiable by any reasonable, educated person without specialist knowledge."

Now back to the subject at hand. I struggled with the bit about assembling primary sources, and I share COgden's concern that simply presenting verifiable facts is not OR, and cherry picking facts is more of a NPOV issue. But I did see the point that it could cross over to an NOR issue if the bias was introducing a new previously unpublished interpretation, rather than just supporting one side in an existing dispute that was already documented by RS. Perhaps I did not address this carefully enough in the draft text, and I think it is worth further discussion, but please discuss it without obscuring the point with the primary/secondary bias. Dhaluza 09:37, 25 September 2007 (UTC)

I already posted this suggestion in the middle of another thread, but the discussion following it skewed off onto other subjects, so I'm posting it in a new section so that (hopefully) the discussion here can focus on the suggestion itself.

Given the history of the section, and comments made by people on both sides of the debate here, it seems that the greatest (potential) value of the PSTS section is to explain what a tertiary source is, and explain that that's what wikipedia is supposed to be. This doesn't mean prohibiting primary sources, but perhaps indicating that they require greater caution. (every definition I've found - and linked during discussions above - of "tertiary source" says that they are based on primary and secondary sources, but mainly secondary ones). As such, I would suggest that the section be redrafted with this clear focus, as that should actually be less confusing and give better guidance to editors. The potential for moving the section to a new page to be referred from any policy that needs it is a seperate matter that, while I still support, is not an aspect of this suggestion. SamBC(talk) 13:19, 24 September 2007 (UTC)

Sam, I gotta go for now, and will be gone for most of the next couple of days. Very quickly, the difference between primary and secondary sources is key, while the difference between secondary and tertiary sources is supplementary to what originally was termed a secondary source, to account for encyclopedias and other compendia derived from secondary and/or primary sources. The key here is 1) to avoid making WP articles a primary source, and 2) to avoid making WP articles a new or original synthesis of primary and/or secondary sources. In other words, as Blueboar succinctly put it, "stick to the sources". Original wording and editorial decisionmaking is one thing, and is permissible, while new, previously unpublished material, and new, previously unpublished analyses or conclusions are forbidden. ... Kenosis 13:49, 24 September 2007 (UTC)

I completely and utterly agree with what you write here (which is a good thing, given the friction we seem to have generated recently - I'll put that down to heightened emotions on both sides). My suggestion above is based on synthesising points that others have made that were, to my mind, cogent and appropriate. It's not an original thought of my own. Whether wikipedia is a tertiary source or not depends on the precise definition, and the definiton isn't as widely agreed (on or off wiki) as primary and secondary. However, wikipedia's intent seems to fit the gist of all the definitions I've found, as what you describe (not adding novel interpretations and so on) seems to be something that' supposed to be a characteristic of tertiary sources, rather than the types of source they use. SamBC(talk) 14:02, 24 September 2007 (UTC)

Although you'll be away a while, I would hope that others either share or understand your concerns enough that we can carry on discussion almost as if you were still here. SamBC(talk) 14:02, 24 September 2007 (UTC)

If we can achieve consensus on what Kenosis suggests, I'm all for it. I don't disagree with anything above. Achieving consensus on this seems a little more ambitious than consensus on something like the proposal Blueboar is managing, which I also support. Kenosis's suggestion has been suggested before, and one of the sticking points was in defining primary and secondary sources. Maybe we don't have to define them in this policy. We just refer to the Wikipedia articles primary source and secondary source. COGDEN 18:12, 24 September 2007 (UTC)

I would define a primary source as any source of facts that are the object of someone's interpretation, explanation, analysis or synthesis, or facts used to interpret, explain, analyze or synthesize other facts. What is wrong with this? Slrubenstein | Talk 18:15, 24 September 2007 (UTC)

Maybe insert "subsequent" in there, as in "...the object of someones subsequent interpretation..."? I had read that twice before it struck me exactly what you saying. At first I thought you were saying that someones interpretations etc. were primary sources. The subtle distinction in there went right by me at first. wbfergus ^Talk 18:25, 24 September 2007 (UTC)

An interpretation etc. can be a primary source, if it's the original source where that interpretation can be found. COGDEN 18:59, 24 September 2007 (UTC)

Michelson-Morley experiment when describing his new theory of relativity, that cited data is a secondary source for that data. But Einstein's new analysis (Relativity) is the primary source of his new theoretical conclusions. If the policy stated that, I wouldn't disagree, but I don't think we need that in the policy. It's not necessary, and doesn't add anything. COGDEN

18:59, 24 September 2007 (UTC)

It was off the top of my head and I certainly appreciate help in phrasing it more clearly and elegantly! But here is why I wouldn't say subseuent: as we all know, peer-reviewed journal articles often publish data and analysis of data simultaneously, and in many disciplines in both the social sciences and natural sciences such articles may be the first time the data is being published - so subsequent analysis may be technically accurate in terms of the process by which such articles get written, but for those of us using such articles to research a Wikipedia article, subsequent may be confusing. Slrubenstein | Talk 18:26, 24 September 2007 (UTC)

Just passing through quickly here, among the things that appear unaccounted in SLRubenstein's proposal is that the works of

Heisenberg and Bohr. Homer's Odyssey and Iliad are examples of primary sources too. More, secondary sources can also be "the object of someone's interpretation, explanation, analysis or synthesis." This needs more work before considering insertion. All, or at least most, of the bases should be covered if such a statement is to be effective. Please don't be too hasty to insert yet. IMO, what's presently needed is careful discussion about how to improve the policy-page description of "secondary sources". For example, textbooks are conspicuously missing from secondary sources, and some kinds of textbooks may be regarded as tertiary sources, particularly introductory textbooks that make use exclusively of established secondary sources. Also presently missing from "primary sources" are nonfiction primary sources-- e.g., I mentioned aseveral philosophers and several physicists above. ... Kenosis

18:35, 24 September 2007 (UTC)

(Unindent)I don't think there would be any major contention on this section, on the basis it is using a description of sources to describe the characteristics of an encyclopaedia. The issue is raised by trying to then reverse the logic (I am trying to think of the right word, is it an intransitive function where you can put a value in and get a predictable value out the other end, but you put the answer back through and you cannot get to the question?). Kenosis recent edit on the main page actually removes that logic and reverts to a general description of sourcing without an attempt to invert the logic, so it becomes far less contentious. Sticking with that thought, if the current description of sourcing as it stands only describes the features of a good encyclopaedia, then we may see that whilst interesting, it does not really underwrite the policing of policy. Spenny 18:41, 24 September 2007 (UTC)

The point of a policy page isn't only to define and aid the policing of the policy, it also has the goal of making the policy understandable. SamBC(talk) 18:44, 24 September 2007 (UTC)

I'm comfortable with that, and I think that is the distinction I wanted to make, but didn't say it clearly enough (the interesting comment wasn't meant to be as dismissive as it sounded). Rather it was an observation that having cleaned up, perhaps we could garner some consensus for the principle that although secondary sources make good encyclopaedias, that does not mean that primary sources make bad encyclopaedias. Spenny 18:51, 24 September 2007 (UTC)

Well... that all depends on how they are used, doesn't it? I know it isn't a popular view right now, but I still come back to the fact that both types of sources can be used correctly and both can be used incorrectly. Blueboar 19:15, 24 September 2007 (UTC)

I would agree with that. I can't recollect for sure (and probably some reading through the recent archives may be required), but it seems like this point has been given examples by people from both sides of the argument, that any source can be misused to insert OR, regardless of primary or secondary.

Now, now... Let's be NPOV here... even Tertiary sources can be misused to insert OR. :>) Blueboar 19:40, 24 September 2007 (UTC)

...and we deal with that by not policing use by these terms, as however you divide the sources, you get back to stick to the sources. Spenny 20:25, 24 September 2007 (UTC)

Actually, just "stick to the sources" is not adequate. WP:NOR also directs the editors to use secondary sources in preference to highly technical or obscure primary-source material. I just gave some examples above, the writings of quantum physicists, leading philosophers, etc. So it is a drective to "stick to the sources", but it's not only to stick to the sources. Where the material is oblique and opinions about it differ, secondary source material is to be used, otherwise it's original research or original synthesis. ... Kenosis 21:20, 24 September 2007 (UTC)

Fair comment, though that is a more complicated knot to unravel. Given that we are writing in summary style, it is that which leads us to choosing appropriate, summarised sources, as we will not find the analysis we seek in the primary documents. The other issue to wrestle with is that is fine when dealing with academia, but how do we evaluate the sources when we have a journalistic editorialisation for the story, not for the facts? As I've commented elsewhere, the likes of Fox News are not reliable sources but more usable sources with a requirement for editorial judgement to be applied when considering them as a source.

Another question is, where we have summary style documents on things like sporting facts, such as league table positions, isn't that basic information safer from the primary sources rather than a journalistic summary piece? If "opinions differ", then in this case are we not simply suggesting editorial incompetence has occurred. Spenny 21:56, 24 September 2007 (UTC)

These issues are expected to be discussed and consensused so as to present a

dovetail among the three. .... Kenosis

22:42, 24 September 2007 (UTC)

OK, if the purpose of defining primary and secondary sources is to explain what Wikipedia is not, then that suggests it be moved to the policy page

WP:NOT. I think that might actually be something that opposing sides could agree is a compromise that could attract consensus. Those who think PSTS is unnecessary obfuscation in NOR should be happy to see it moved, while those who think it is essential policy for WP in general will see it remain in a policy page. The history shows that this section was originally inserted to state that WP is a secondary source, but that changed over time to show it is a tertiary source; so now that this is more about what WP is not, rather than what it is, perhaps a rethink is in order.Dhaluza

09:24, 25 September 2007 (UTC)

Kenosis is right that

WP:ATT neatly avoided this whole argument about which policy this goes in, and is a worthwhile approach. However, in the current framework PSTS is particularly related to the need to ensure that primary sources relative to the article's subject are not used to create any interpretation or synthesis, which must come from a secondary source and be properly summarised and attributed. Either kind of source can be misused and should not be misused, but for NOR the essential point is getting the interpretation etc. from an appropriate source. .. dave souza, talk

09:58, 25 September 2007 (UTC)

Transclusion example

Okay, here's an example of transclusion User:Wbfergus/Sandbox/Transclusion example. I copied the "Sources section from the policy to create the page. I next created a sub-page called "Sources", and moved the PSTS sub-section into that sub-page, then I used the transclusion template on the first page to pull in the sub-page. Feel free to look at the edit text to see how it works.

Now, the advantage to this is that if we implement this on this policy, the "Sources" can be a sub-page and any discussion about these "Sources" can be discussed there, instead of here. This would help focus this talk page more to strictly NOR issues, and all of the "Sources" discussions happen somewhere else. When reading, it is transparent, so it appears that what is transcluded is hard-coded into the text they see on the screen, so there is no "loss of policy" or other weakening of this policy, it's merely moving one section that has generated an awful lot of discussion into another spot so thse discussions can happen elsewhere.

A disadvantage would be that it would need to be documented better what is happening and that "Sources" discussions happen on the sub-page. This isn't that big of a deal, but it's a step that needs to be done. (I won't sign the two questions below so that my sig doesn't interfere with the questions) wbfergus ^Talk 20:02, 24 September 2007 (UTC)

Does anybody see any other advantages or disadvantages?

For the record, I'm experimenting (on the above example page) with a template to show how an "announcement" about where to conduct "Sources" type discussions could be incorporated at the top of this page. I'm not very good at the template syntax, so it still needs a few tweaks if anybody cares to have a go at it. wbfergus ^Talk 22:43, 24 September 2007 (UTC)

Would all parties involved agree that this would be an effective compromise to at least "clean-up" this talk page so that it is far easier to concentrate on solely NOR issues?

I don't see any benefit to transclusion, and lots of potential for confusion. If a full cost-benefit analysis of various types of sources is going to be part of the policy (which seems unlikely at this point), then it should just be included. If it's moved to an essay, it should be referenced by a link. COGDEN 20:17, 24 September 2007 (UTC)

The sub-page isn't an essay. Being a sub-page, it is still part of the present policy, just located a level beneath. So instead of the text physically residing within the NOR page, it is located a level below at NOR\Sources instead. I "Think" the only real advantage to this would be to get all of the talk about "Sources" off of this talk page and onto it's own "specialized" talk page, so that it is far easier to look at this talk page and only see NOR related discussions instead of having to browse through several hundred thousand KB of text to see if there are any "non-sourcing" type discussions. It would allow all of the discussions about sourcing problems or issues on it's own page. I may be wrong though, maybe people like having to wade through reams of electrons to see if there's anything new or pertinent. wbfergus ^Talk 20:32, 24 September 2007 (UTC)

There are many instances of policy pages that reference essays. That doesn't incorporate the essay into the policy. It just gives the reader some place to go for a discussion on that topic. We could do something similar here. COGDEN 20:38, 24 September 2007 (UTC)

Who is talking about an essay? 21:42, 24 September 2007 (UTC) —Preceding unsigned comment added by Jossi (talk • contribs)

I think that it has been stated by several people in previous discussions above that moving PSTS to a guideline or especially an essay is unacceptable. The only way they would agree to moving it would be to a policy page, either a brand new one or an existing one. At least with transclusion, the contentious section is moved, but not out of the policy until such time as everybody can reach some agreement on where. This way, it is still part of this policy, so the policy remains unchanged, but we can separate the talk pages into a NOR talk page and a talk page for the "Sources" discussions. It's not much of a change, but it is some change. I see it as a small step forward, I don't know how others see it. wbfergus ^Talk 22:08, 24 September 2007 (UTC)

If it's moved, and then described as a "policy page", it would promptly get demoted, since it does not currently have consensus. I see transclusion as a step sideways, rather than a step forward. I'd rather focus on solving the problem, which is that the current policy has a non-consensus section. COGDEN 22:19, 24 September 2007 (UTC)

Well, there no doubt the section iteself is contentious. It's been that way since at least January 2005, but the arguments are still going on with little to no improvement. This way, those who argue that it still needs to be part of this policy can remain happy that it is still part of this policy. For those who at least want this talk page cleaned up a bit, so it's easier to follow pertinent 'stuff' about NOR, then they can be happy as well. For the remainder, that want to completely get rid of the PSTS section, well, maybe this could be a first step. Maybe if these discussion center around NOR issues, it may be apparent that PSTS can live on it's own somewhere without detracting from this policy, I don't know. At least it's a small step forward that I think can keep everybody happy. I'm all for cleaning up this talk page, and not just by archiving. wbfergus ^Talk 22:40, 24 September 2007 (UTC)

Would "demoting" the PSTS material to RS really make it any less stable or effective? I don't see that it would. Even here it's not a hard-and-fast rule and there are exceptions. Anyone who wants to make an argument or an article revision based on sourcing could still do so and simply cite RS instead of NOR - people revise articles all the time based on guidelines or even essays. There are a number of people who carefully watch RS and would resist any ill-considered changes. Finally, I'm not saying this is what we should do but one alternative is to promote RS to policy status in connection with a move. There's already a universal consensus that we need to use reliable sources, and there's little if anything in the RS guideline that's controversial at all. In fact, there's little substantive material on that page at all. It's mostly a collection of links and brief summaries of various other pages. It's a nearly empty framework that's ready for some content. If we get consensus both here and on RS for promoting RS to policy, any demotion back to guideline could be summarily reverted. Wikidemo 22:54, 24 September 2007 (UTC)

There is no rush to transclude. ≈ jossi ≈ (talk) 23:21, 24 September 2007 (UTC)

Also note that we do not need RS to become a policy page, as the reliability of sources is already explored in WP:V, WP:NOR and WP:NPOV in their specific contexts. ≈ jossi ≈ (talk) 23:23, 24 September 2007 (UTC)

True, I don't see a rush either. Just an option. In fact, I don't see much of a rush to do anything here, and the way things have been going any quick action seems unlikely. There's a question of organization, quite apart from those challenging the appropriateness and wording of PSTS: is it better to have RS information reside in these various policies, or centrally (or transcluded, or duplicated, two less attractive options IMO).

The only "rush" I see is that this page is already 468KB long, and past due for archiving, though there are still lots of active discussions going on. This will also be at least the third archive devoted almost entirley to "source-type" discussions. If the transclusion is acceptable, and others think archiving is past due also, then it might make sense to try transclusion for a while and see if it does help clean up this talk page. It can always be reverted back fairly simply if it doesn't work. wbfergus ^Talk 00:58, 25 September 2007 (UTC)

A large part of the confusion between source types seems to lie in contradictory understandings of the distinctions. For example, my main area of study has been in the humanities, which uses quite different definitions than the sciences, which use differing definitions than the "strict dictionary sense". It seems like a lot of the confusion and disagreement largely lies in those distinct understandings of the issue.

I will raise peer-reviewed publications for an example. In the humanities, peer-reviewed publications are considered secondary sources. In the sciences, they are generally considered primary sources (when putting forth a novel theory). In the strict dictionary sense, they are a mixture of primary and secondary, based on identifying raw data and novel claims as primary material. Within these individual fields and viewpoints, there are further variations as well. For example, a small part of the humanities considers any source drawn upon in research to be a primary source. Beyond all this confusion, the academic distinctions were created with the goal of creating and reviewing original research in mind.

In a broad sense, primary sources are those drawn upon as "raw" evidence, which are subjected to an intense review and analysis, particularly of their reliability and context. (In humanities, it's more "internal" to the work, peer-review generally judging the strength of argument. In the sciences, it's more "external", peer-review usually examining even the raw data collected with intense scrutiny.) In the same broad sense, secondary sources are the analysis and reviews of other contemporary authors, drawn upon to critique and counter those views or cited as sympathetic to the central thesis being forwarded.

So, in essence, I feel the conflicting definitions create a lot of confusion and conflict, and I have no particular ideas of how to "fix" this problem. A centralized definition is not a solution in and of itself, because such a thing has been present in policy, but the confusion and conflict continue. Perhaps we need to work with other terms, or maybe the language simply needs to be clarified and tightened. I also wonder if definitions created to promote and frame original research are appropriate for a policy that has the opposing intention. We may be working at cross-purposes to ourselves in continuing to use definitions created for the converse purpose as the policy they support. Just some thoughts. What are your thoughts in response? Vassyana 20:24, 24 September 2007 (UTC)

Although you may be correct that in humanities, peer-reviewed articles are often considered secondary sources, I think that's just the result of imprecise terminology. For undergraduate purposes, you might not care so much about the fine distinction. But in academic historiography, where the terms originated, the terminology is more precise: Primary sources include original historical ideas, whether or not based on underlying data (the original source of the underlying data, of course, would be primary too). Secondary sources are just historical ideas as repeated by others (not the original author of the idea). The terms are used the same ways in works regarding the history of science and technical fields.

I think this complexity and confusion can be avoided by simply not going there. We don't need the distinction in order to express what OR is. OR can exist when you cite any type of source, and we have all sorts of inconsistent views on which which type of source is "better". But we don't need to go there. Let's just focus on the core issues: (1) how do you cite a fact without committing OR, and (2) how do you cite an analysis/conclusion/etc. without committing OR. We don't need fancy high-falutin distinctions between primary/secondary, etc. COGDEN 20:53, 24 September 2007 (UTC)

Cogden, seeing as pretty much everyone else involved disagrees with you on this interpretation, do you have any references that clearly agree with that interpretation of the primary/secondary distinction? SamBC(talk) 21:32, 24 September 2007 (UTC)

You obviously are severely misunderstanding, or simply unaware, of how things are handled in historiography. I'd recommend An Introduction to the Historiography of Science published by Cambridge University Press, as a good start. It will help you better grok how things are done in historiography and the history of science, which are quite contrary to the points you are putting forward. Vassyana 22:47, 24 September 2007 (UTC)

See below. My understanding is consistent with the history/historiographic texts I cite below. This shows you how confusing the issue is, and how inappropriate these fine distinctions are to a Wikipedia policy page. COGDEN 23:58, 24 September 2007 (UTC)

I don't disagree that the subject is confusing, with many definitions across many fields. You're using the library science definition of primary sources (which is often used in ENG-XXX lit classes), not the history or historiography definitions. However, that does help me understand where you're coming from a bit more. Vassyana 00:23, 25 September 2007 (UTC)

I think the problem with source typing using primary and secondary, and especially its inclusion in NOR, is that it does not meet the standard of being easily verifiable without specialist knowledge. Dhaluza 09:10, 25 September 2007 (UTC)

Vassyana, in the humanities secondary sources such as peer-reviewed journal pulbications can be sources of data and thus be used as primary sources just as in the physical sciences. For example, a historian my publish material from an archive in an article, along with analysis and interpretation; another historian can use that material to promote another interpretation in another article. Someone studying medieval French literature can publish an article with a hithertoo unpublished poem, and analyze it - and another scholar can then draw on the article for the poem, and provide another interpretation. I do not see how this is different from an article on molecular genetics or chemistry providing data. The point is, data can be published independently or in the context of a peer-reviewed journal article, along with analysis or interpretation. This is true for the natural sciences and the humanities. Slrubenstein | Talk 09:58, 25 September 2007 (UTC)

It would be poor scholarship to draw primary data from secondary sources in that field. For example, you shouldn't use the photographs of the Dead Sea Scrolls published in the first article they appear in as a primary source. You should acquire access to the photographs (or authorized high-quality reproductions) themselves. The only time you should draw upon primary sources from secondary sources is when specifically writing about how that author, or a group they belong to, used those sources. The results of a dig and an article about the results of a dig are quite distinct. It is my understanding that in the natural sciences, there is not such a separation of primary and secondary. As far as the humanities, if we treat it as they do in the field, additional limitations on the use of primary sources appearing in secondary sources is not necessary (and on the contrary, cumbersome). That is, if the primary materials from those secondary sources were only cited in the context, including accompanying analysis/interpretation, that they appear in the secondary source. That would be purely relying on a published secondary source, not subject to any restrictions about using primary materials. The humanities generally do not present "raw data" absent immediate context as the natural sciences do, except when making a publication of purely "raw data". For example, the publication and distribution of the Dead Sea Scrolls images alone (with minimal commentary for identification purposes and minor "technical" notes about obscure characters/damaged portions) would be a primary source. A modern peer-reviewed article using those photos to form part of the evidence supporting the central thesis is not in any way considered a primary source, within history/anthropology. Vassyana 14:21, 25 September 2007 (UTC)

I'm also interested in your thoughts regarding using definitions intended to aid original research for the purpose of disallowing original research. To make a crude analogy, one would not use brakes to accelerate. So, why should we use distinctions intended to promote original research in this policy? Additionally, how can we use any particular definition when various fields of scholarship treat them in very distinct fashions? What is the purpose and intent of the distinctions in this policy? Vassyana 14:42, 25 September 2007 (UTC)

Well, to be fair, most peer-reviewed journal articles in the life or physical sciences never present raw data. What may look to you like raw data is often actually the rsult of several stages of processing. My point is simply this: in both works from the natural sciences and humanities - works that make an argument, that interpret or explain data - there is often data, in some form, that others can use to develop their own interpretations or explanations. I very strongly believe in making a distinction between data and arguments about data, and I think that the distinction between primary and secondary sources is useful, I am just saying that secondary sources can be sources of data, and this is true across academe. I am not sure about your second question - I do not understand the question. I think your analogy is false. In the United States there is a rule about how old people must be to drink; I think in all States it is now 21 years old. The difference between being twenty years old and twenty one years old is important for people in both groups, but for the first groups it means they must buy a fake ID to by booze, and for people in the second group it means they do not. for people who sell booze, it means they have to check IDs and know how to tell the difference. Knowing the difference has opposite effects for two groups. There are all sorts of distinctions in life that allow one group of people to do something, and prohibit others from doing the same thing. There is nothing unusual about this. Slrubenstein | Talk 16:42, 25 September 2007 (UTC)

Thanks for the response. I understand your analogy and it makes sense. I think if we are to keep the distinctions, we should be careful to bear in mind they were created to promote original research to avoid inadvertently doing the same or causing confusion. The origin and purpose of terms is important in determining their future use and modification. My second question may have been poorly worded. In essence, how can we avoid confusion over the definitions, or appeal to "the" academic definition, when there is a large amount of confusion/disagreement over the use of the terms between various fields? For example, peer-reviewed articles are always secondary sources in history and anthropology, regardless of whether they contain "data" or make novel assertions. In the same field, historical sources are always primary sources. In historiography and many branches of literature, a primary source is exclusive to the period or style being studied. In library sciences and other branches of literature, "primary source" is used very literally to indicate the primary source of a statement, claim, et al. There are a lot of conflicting definitions and meanings, and I think this is a large part of what contributes to the confusion here. Basically, we need to acknowledge these several conflicting definitions and what fields they occur in, at least when we're addressing PSTS and the need (or lack thereof) to revise, rewrite or remove it. I hope that better explains the intent of my second question. Vassyana 18:19, 25 September 2007 (UTC)

This is much clearer, and I agree with everything you say. I think we just need to define terms clearly, and also emphasize that what is important is how material is being used - and for that reason, terms can apply to overlapping categories. Most peer-reviewed journal articles in the life and physical sciences do not limit themselves to raw-data and easily fall under our definition of "secondary source." The point is simply that secondary sources can also be sources of primary material (data), and I don't think it should be hard to make this clear in the article. It seems like people view primary/secondary/tertiary sources as mutually exclusive categories when I think they are either overlapping or even best understood by degrees of inclusion - tertiary sources refer to secondary sources and primary sources; secondary sources refer to primary sources ... at this point since we both agree this discussion should move to the following section. I recently posted a long comment calling attention to three distinctions we need to make in this discussion - perhaps you can comment, amplifying on or correcting what I wrote there. Slrubenstein | Talk 18:28, 25 September 2007 (UTC)

I posted much of this before, but it wasn't here very long before being rotated to the archive pages. This is a survey of good sources defining primary and secondary sources. I don't think that this distinction needs to be in this policy page, and it overly complicates things, but here it is anyway, so that we can avoid making statements in the policy that disagree with common usage.

Definitions of a primary source:

• "Manuscripts, records, or documents providing original research or documentation." [4]
• "Primary sources contain original information and are usually the place where the original information first appears. Examples of primary sources include interviews, diaries, letters, speeches, results of experiments or original research, literary works, autobiographies, original theories, and other materials. Compare to secondary source."[5]
• "An original work such as a book, manuscript, or document produced by an author. A book can also be a secondary source. A scientific or scholarly journal article is usually a primary source."[6]
• "The main source used to defend a research question. For example, critical essays, documented studies, scholarly or technical journals, or interviews with experts."[7]
• "Information which has not been interpreted by another person, ie, original articulation of an idea or concept."[8]

Definitions of a secondary source:

• "second-hand report or review of original research that is written by someone other than the original researcher"[9]
• "A secondary source is a record or statement of an event or circumstance made by a non-eyewitness or by someone not closely connected with the event or circumstances, recorded or stated verbally either at or sometime after the event, or by an eye-witness at a time after the event when the fallibility of memory is an important factor." Derek Harland, A Basic Course in Genealogy: Volume two, Research Procedure and Evaluation of Evidence, (Bookcraft Inc, 1958)
• "A secondary historical source is a report about the past which does not have authority. A book by a scholar about history is a secondary source. Thus, González' textbook Story of Christianity is a secondary source. You can't prove anything about the past from a secondary source." [10]
• "The distinction between primary and secondary sources is only meaningful when applied to symbolic sources. Moreover, the distinction is not a sharp one. Since a source is only a source in a specific historical context, the same source object can be both a primary or secondary source according to what it is used for. Duhem's la theorie physique will be a useful secondary source for the historian who wishes to study the history of the theries of gravitation; it will be a fine primary source for the historian who wishes to investigate positivist views of science at the turn of the century." [11]
• "A primary source is a document in which an event is described by its witnesses or first recorders, for example: The Diaries of LouisRiel (McL FC3217.1 R54A3). It is the raw material or data which the historian must evaluate in studying the history of a period, event, or individual. The historian produces a secondary source based on primary sources, e.g. Prairie Fire: the 1885 North-WestRebellion by B.Beal and R.Macleod (McL FC3215 B42 1984). The concept is relative: any document may be used as a primary source. For example, if you were to study the historiography of the Riel Rebellion, the book by Beal and Macleod may be used as a primary source." [12]

There are, of course, sloppier definitions of secondary sources. Some say that a secondary source is something that comments on a primary source. And yes, that's very true: if something comments on a primary source or contains analysis about prior materials, it's definitely a secondary source. But if the commentary is original research, it's also a primary source as to that new idea. The terms are relative, and it all depends on what you are focusing on, and sources can be both primary and secondary. But why do we even need to even go there? The concepts of primary and secondary are confusing, and not always interpreted consistently. The ambiguity between primary and secondary sources is all very complicated, and this whole area really has no place in a policy article addressed to lay editors. This is stuff that PhDs debate about. How can we expect the average Wikipedia editor to understand and apply this? COGDEN 23:55, 24 September 2007 (UTC)

The second and third links under primary sources are dead (404 errors). For reference, those represent the library science definition of primary sources. Vassyana 00:30, 25 September 2007 (UTC)

IMO, it is quite arguable that the wikilinks to articles on

tertiary sources should not be on the policy page in the first place. They were added on February 1, 2007 here. The problem is that when a policy page wikilinks to a main namespace article, the article may be dealing with different contexts, and in this case they do just that. Two different standards. These should instead link to project pages, not articles. Nonetheless, there are fairly standard usages of these terms by libraries and historiographers and others. If participants wish to keep the wikilinks to the articles, at the very least the articles should be capable of reasonably accommodating the policy-page interpretations of PSTS among the usages described in the article. The convention in WP, though, is that policy concerns do not wikilink to articles to define their terms. Alternately, Dave Souza and others seem to like the idea of a separate WP:PSTS project page, in which case the examples and discussions of PSTS can of course be extended to a more comprehensive treatment. ... Kenosis

02:16, 25 September 2007 (UTC)

A policy page should not link to the article space for definitions for the reasons I stated above. It is not good for either page. People will defend the article like a policy, and if the article does change to take into account new information, it can change the policy in unintended ways. Dhaluza 09:04, 25 September 2007 (UTC)

The Wikipedia convention is to not redirect pages across namespaces. For wikilinks, the convention is not as strict, but as a practical matter, the wikilinks to articles was a simple oversight. More, these three articles were not closely attended to or particularly well written, and the one on tertiary sources has been almost completely neglected since its inception well over two years ago. The standards for articles are different than they are for policy pages. I've removed the wikilinks. If Dave Souza's proposal to start a separate WP:PSTS page gets going, ultimately these can link to appropirate places on an appropriate project page in the future. ... Kenosis 13:38, 25 September 2007 (UTC).

One thing I have realized in working on the drafts is that the definition of "primary source" does not distinguish between raw evidence, and the first analysis of that evidence. The two are very different, and need to be handled differently, but the arbitrary definitions of primary and secondary don't allow for this. If we treat raw evidence as the true primary source, then an initial analysis needs to be a primary-and-a-half source to allow the analysis of the analysis to be secondary. Dhaluza 09:01, 25 September 2007 (UTC)

Yes, some of these things are readily debatable, and to a logical positivist they may be intererably inexact. The following link at the University of Maryland library, though, is illustrative of the general thinking among library-sciences specialists: here. The UMD library is only one of a significant number of post-secondary libraries that use this same approach. Note how "diaries" are primary sources, books on the historic event are secondary sources, and encyclopedias on the topic are tertiary sources, according to that method of classification. The divisions are, of course, not hard-and-fast, so it will inevitably make strict interpretationists crazy. But note that many widely accepted editorial considerations make strict interpretationists crazy (which may be partly why technical authors in the "real world" are sometimes tempted to come into meetings with non-technically oriented publishing editors with a holstered weapon and the holster strap already unsnapped ;-\ and also why the editor keeps one in the top desk drawer too ;-) . So I wouldn't make too much of the fact that the delineations aren't hard-and-fast, because WP editors are also expected to use judgment and mutual cooperation in the process of writing articles. ... Kenosis 15:22, 25 September 2007 (UTC)

Library scientists are great people, but they are not who we should be listening to here. We should listen to historiographers. Library science has to do with classifying books and periodicals, and their definition is rough-and-true, but not as sophisticated or correct as the one used by historiographers. Historiography has to do with tracing the history and sequence of ideas. Original research is about new ideas, and at least when the subject matter of the article is history, the OR question comes down to one of historiography.

Moreover, even the particular definition you cite blanket-classifies interviews, conference proceedings, and works of literature into a as primary sources. These are the bread and butter of Wikipedia articles, and shouldn't ever be discouraged. But, once again, if there are different definitions of primary and secondary sources, why are we basing our policy on such definitions. We don't need them! And we don't need the confusion and ambiguity! COGDEN 17:45, 25 September 2007 (UTC)

The opinion is duly noted. Wikipedia has its own policies that are not based upon editorial policies as defined in other places (except perhaps in guidelines such as

WP:MOS). And similar issues occur in many policy discussions in Wikipedia. See, for example, the usages of "free content" and "non-free content in WP:Non-free content, which bear little resemblance to the definitions of "free" in most available external sources, and where public domain technically falls under "non-free". In other words, Wikipedia has been entitled to define its own terms of use, and it has indeed done so from the beginning. ... Kenosis

17:48, 25 September 2007 (UTC)

WP:Non-free content#Applied to Wikipedia explicitly contradicts your assertion. Vassyana 18:25, 25 September 2007 (UTC)

Pardon me, but no, the section referring to copyright delineations is only one aspect of

WP:PSTS, it already reasonably well aligns with other uses such as "library science" and even historiography to some extent, but is not obliged by any stretch of imagination to precisely track a particular usage of the terms as used in another field. ... Kenosis

18:44, 25 September 2007 (UTC)

You assert that public domain is not equivalent to free. However, the section I linked explicitly says: "If a work has no copyright or is licensed to Wikipedia under an acceptable "free" license, it is a free work and may be used on Wikipedia without copyright concerns. See public domain, copyright, and Cornell University's guide to copyright terms for discussion of works that are not covered by copyright." [emphasis added] It quite clearly and explicitly has public domain works as free works. It's not an opinion, it's plain reading comprehension. Vassyana 19:40, 25 September 2007 (UTC)

The presumption here that "public domain" is equivalent to WP's definition of "free" is incorrect. A free license is an affirmative act, explicitly granted by placing a statement in conjunction with posting the image. Public domain normally occurs by default and can be very difficult to ascertain in many instances. So they're two very different things. In Wikipedia, "public domain" is in something of a "black hole" w.r.t. "free" and "non-free". And Wikipedia most certainly does not explicitly mirror copyright law. The NFCC are far more strongly defined than copyright law. In other words, WP can, presently does, and has from the beginning set its own criteria. Also, if you look at the history, you will see that I was the user who added the material you just referred to at WP:NFC here. In other words, I'm not in opposition to external references in policy and guideline project pages. But, Wikipedia is not bound by such external considerations, except as regards compliance with applicable law. Nonetheless, as I also indicated, WP:PSTS reasonably tracks usage of PSTS in the external sources. .... Kenosis 01:42, 26 September 2007 (UTC)

(outdent) I'm in no mood to play rhetorical games, so I'm going to be painfully blunt. You should bother to read and pay attention. The section you link is a different section that the one I pointed out. What exactly is unclear to you about "[i]f a work has no copyright ... it is a free work"? If a work has no copyright (like works in the public domain), it's considered a free work. That's pretty damn clear English. Vassyana 01:55, 26 September 2007 (UTC)

These, Vassyana, are by no means rhetorical games, nor do I feel at all pained by your bluntness. I do now see the quote you refered to-- when I saw the Cornell link in the first link you provided I assumed it was the part I added, which also contained the Cornell link then under very active discussion in that explicit context, in addition to several more I provided. The specific passage with the statement referred to by you (Vassyana) was also added on the same day as the passages and copyright-research links I added to the guideline portion of the WP:NFC page, same section, different subsection (both 10 September 2007, the one you referred to was placed by Wikidemo here). And it was added in significant part in response to my repeated explanations about the neglect of public domain in the NFCC accounting. And, the difference between an explicit free license, which is explicitly granted in black and white, so to speak, and public domain, which occurs by default which is sometimes excruciatingly difficult to verify under modern copyright law, is by no stretch of imagination a "rhetorical game". The difference between the two still confounds WP policy, and is what held up even the guideline page on that particular matter. Thus, as I was saying, I am not opposed to referencing external usages of Wikipedia terms in either policy or guideline discussions. And both of the statements you (Vassyana) and I just referred to are in the guideline portion of WP:NFC. "Public domain" is still, as I just said, in a "black hole" w.r.t. how WP policy (the NFCC) presently analyzes "free" and "non-free". Presently there is no clearly defined policy on public domain images and other public domain media. That is not semantics or rhetoric. Nor do I consider this rhetorical games, or I wouldn't bother to attempt to put these issues into perspective for you.

The fact remains, nonetheless, that Wikipedia is not, as is being repetively argued on this page w.r.t. WP:NOR, required to mirror the external usages of its policy terms. I just pointed out another arena of policy where Wikipedia chooses its own terms, and it's not limited to

WP:BLP. These issues too are by no means "semantic" or "rhetorical". Some of the recent rapid-fire arguments by opponents of WP:PSTS appear hell-bent on picking holes in every little thing, with one objective, and only one objective -- to change longstanding Wikipedia NOR policy. In my opinion, if anything ends up being a rhetorical game, it is this ongoing struggle by a few WP participants to pick holes in longstanding, very widely consensused policy upon which many thousands of users have come to rely for quite some time now, without a coherent replacement plan that participants can agree is an improvement. Presently, in this section, the argument is that it does not mirror external usages of PSTS. In other sections it's either one thing or another, or some combination of complaints, but always without a coherent plan that has potential to be widely agreed upon as a superior replacement. Well, the plain fact is that WP:PSTS is not exact, but WP's usage of PSTS comes damned close to that of library science usage and also to that of historiographers. Again, as I was attempting to point out, even if WP:PSTS did not closely mirror certain external usages of PSTS, WP is still within its rights to use its own standard. Note also that the entire policy WP:NOR does not mirror usages outside of Wikipedia, and has only vague analogues with other encyclopedia writing and journalistic writing. Not that I expect these observations to bring a quick end to these various attempts to pick holes in whatever appears like it can be argued. But I offer it just for your information. ... Kenosis

03:18, 26 September 2007 (UTC)

Thank you for the well-considered and polite response, despite my terse and tense message to you. It is appreciated and makes sense for the most part. However, I would ask you to reconsider the broad brush you paint with, in regards to those taking some issue with the current PSTS section, as I feel it is grossly inaccurate. Vassyana 04:50, 26 September 2007 (UTC)

No problem, Vassyana. I think most of us can agree that there is still more work to be done on WP:PSTS. In my view, it's still missing a number of important examples that can be stated in one or two words, hopefully without significant controversy. I believe I mentioned religious scripture a bit farther below as needing to be in "Primary sources" -- just one of a number of things that can very readily bring this project page more in keeping with how it's actually being used around the wiki. ... Kenosis 05:06, 26 September 2007 (UTC)

[Posted prior to the above]Again, though, the simple breakdown can be misleading if taken too strictly. For example, in the case of seminal works, whether fiction, non-fiction or scientific, the seminal books themselves can be primary sources. Examples I gave above include the works of

Heisenberg and Bohr are primary sources, whether or not they drew on others' experiments. Homer's Odyssey and Iliad

are examples of primary sources too, at least from the standpoint of Wikipedia editing. Thus, some flexibility and common sense will need to be exercised both here on the policy page and on the article pages. The key in the primary/secondary/tertiary source categorization is not necessarily which record is closest to the ground, so to speak. To be technical, or tendentious, one could argue, for instance, that only the first laboratory notes are primary sources, so by this measure by the time we get to writing in Wikipedia, the textbook source that mentions the experimental result may be a quinary (5th order), senary (6th order), or even farther down the chain of interpretation. For our purposes, generally speaking, the highly technical interpretation by the original developer(s) of a theory is a primary source, and the explanation of the summary results is a secondary source.

In a similar vein, Immanuel Kant's Critique of Pure Reason, a leading turn-of-the-19th-century philosophical treatise that gave us, among other things, the word phenomenon, is a primary source. Since it is a highly obscure work, scholarly interpretations of Kant's work are secondary sources on which we depend very heavily. Several times we've encountered editors who think they know Kant better than the scholars, and that's original research -- which is another example of why PSTS is part of the policy page. Thus, again, these are not hard and fast delineations in WP:PSTS, but they are critical in the implementation of the NOR policy. . ... Kenosis 15:47, 25 September 2007 (UTC)

The very difficulties and ambiguities you describe above show why this policy should not be based on the primary/secondary model. You end up arguing about what is primary and what is secondary, and you lose focus on whether or not what you are trying to do is original research.

As to Kant, while I'm sure editors think they know philosophy better than Kant, they do not know what Kant wrote better than Kant. Thus, if you want the article to say what Kant wrote, you quote Kant. What Kant wrote is a simple factual matter. If, on the other hand, you want to say what Kant meant, then you have to cite either Kant (if he himself said what he meant), or another author. The primariness or secondariness of Kant doesn't matter. What matters is that the editor doesn't say something that hasn't been said before, either by Kant or by another author. COGDEN 17:51, 25 September 2007 (UTC)

<undent> The list of definitions is good, and would make a useful addition to the Wikipedia:Classification of sources essay which has the potential to be developed into the shared PSTS resource being suggested. As a non-academic I'm comfortable with them all and with the essential point in NOR at present that "Primary sources are documents or people very close to the situation being written about.", which means to me that the categorisation is relative to the subject of the article or section. Linking to articles as definitions is undesirable and I agree that it should not be done, which is why there's a place for these links to be replaced by the Classification of sources essay once it's developed into a guideline. Alternatively, the clarified classification could become part of WP:RS .. dave souza, talk 09:50, 25 September 2007 (UTC)

Three points which I think are crucial for a productive discussion. First, we need to distinguish between definitions and uses; what something is and how it is used are two different things. In some cases we do need clear definitions of terms. In other cases, how something is used may be more significant than what exactly it is. Second, we definitely need to distinguish between definitions and examples - I appreciate COGDEN's efforts above, but he mixes up the two and it would help to keep them separate - especially because it is possible that, if we identify something by its use, something can be an example of two different classes (primary vs. secondary or whatever words we choose to use); a house can be a place to live for one person, and a source of income for another. Finally, I would suggest that in some if not many cases, something may be both a primary and a secondary source, depending on how it is used. Not only may one person's primary source be another person's secondary source, but the status or use of something may for most people change over time. I hate to bring up a possibly contentious example, but most modern scholars believe that the Hebrew Bible was itself based on a variety of sources, virtually all of which were lost to its readers by the time it was canonized (some of these sources have been discovered thanks to advances in archeology in the 19th century - but most of it sources are presumably still undescovered, perhaps lost). It is quite likely that when the book of Genesis or Kings was furst composed, everyone viewed it as what we might call a secondary source in the way Gibbon's Decline and Fall of the Roman Empire is (or was!) a secondary source; a synthesis drawn on other texts with a particuloar point of view. Today it is a primary source and the object of considerable analysis and interpretation. I do not propose these distinctions to make our task harder. On the contrary, I think if we bear these in mind we can avoid some muddles which will help us focus on the more difficult challenges. Slrubenstein | Talk 16:54, 25 September 2007 (UTC)

Religious scripture is a poster child for primary-source material, and dealing with it is a case where secondary and tertiary sources are key in Wikipedia. See, for just one example, the approach in

WP:NPOV#Undue_weight, the reliability of which should be assessed in keeping with newer scholarship based upon newer primary sources such as literature on archeological discoveries and any other more recent evidence discussed by scholars, and in more recent secondary sources. Gibbon's book is a primary source only if the subject of the article focuses on Gibbon's book. ... Kenosis

17:36, 25 September 2007 (UTC)

I think the points you raise are well-thought and coherent. I'd also agree with Kenosis that religious scripture is a poster child for primary-source abuse. I am uncomfortable with the PSTS terms, because of confusion and disagreement. However, I would be more comfortable and supportive of a PSTS section that clearly addresses the points raised by Slrubenstuin immediately above that is crafted taking into account my concerns about the conflicting/definitions and origins of the terms in the preceding section. Vassyana 18:40, 25 September 2007 (UTC)

Since I do not disagree with anything Kenosis just wrote, your comment seems a bit like a non-sequitor ... which makes me wonder whether you (Kenosis) misunderstood what I wrote. Slrubenstein | Talk 18:16, 25 September 2007 (UTC)

I agree with the above. My view is that given the difficulties and ambiguities in definition you've just described, we should run the other way screaming. Policy pages are supposed to document and simplify things, not make them more complicated. We don't need people arguing about what is a primary or secondary source, when they should really be arguing about whether the editor's use of the source constitutes original research. Whether the Bible, or whatever, is primary or secondary makes no difference. The same standard applies either way.

For the record, though, I think that if we use the primary/secondary model at all, the only feasible option would be to base the model on use. Reliable sources are neither good or bad in themselves. It's how they are used that can cause problems with OR. But unfortunately, the "use-based" primary/secondary model is also the most complicated one, and the least likely to be understood by lay editors. COGDEN 18:03, 25 September 2007 (UTC)

They're not ambiguities to any unworkable extent. Key is how one views the PSTS breakdown. Primary sources for WP articles are not necessarily the first stage of reporting of facts (though they can be). Primary sources are the reliable publications closest to the origin of the topic of discussion. Secondary sources are published materials and books that are built on primary sources and which analyze and synthesize them. Tertiary sources are, in general, broader syntheses and summaries such as encylopedias, compendia with internal commentary and, arguably, introductory textbooks. ... Kenosis 18:08, 25 September 2007 (UTC)

Per discussion, I revised the introduction, taking into account coments and edits by other people active here.

I would not like to propose to change the following section, "What is excluded." I have three problems with this. first, it opens up not with a statement about what is excluded, but with the motivation behind the policy, so the title of the section and part of the contents are inconsistent. Second, the list of what is excluded seems ad hoc. Third, following sections explain in greater detail what is excluded and why.

Here is what I propose: since we are still debating primary versus secondary sources and what is actually included and excluded, I suggest that instead of calling this section "What is excluded" let's follow the lead of the first sentence and make it a section about what motivated the policy i.e. its origins. Such a section can discuss things that have traditionally been excluded, but in the context of the history of the proposal. I think this explanation of the history would be educational.

ALL I am proposing right now is to change the second section. I am not proposing anything about the third or following sections (on sources and synthesis).

here is what I propose:

---

Origins of the policy

The core policy of Wikipedia,

is meant to provide a framework whereby editors with diverse, often conflicting, even opposing points of view can collaborate on the creation of an encyclopedia. It does so through the principle that while it is often hard for people to agree as to what is the truth, it is much easier for people to agree as to what they and others believe to be the truth. Therefore, Wikipedia does not use "truth" as a criteria for inclusion. Instead, it aims to account for different, notable views of the truth. First codified in February 2002, the objective of the NPOV policy is to produce an unbiased encyclopedia.

In the year that followed a good deal of conflict on article talk pages involved accusations that editors were violating NPOV, and it became clear that this policy, which provided a philosophical foundation for Wikipedia, needed to be supplemented. Wikipedians developed the concept of "

" as a way of ensuring the accuracy of articles by encouraging editors to cite sources; this concept was established as a policy in August 2003. Verifiability was also promoted as a way to ensure that notable views would be represented, under the assumption that the most notable views were easiest to document with sources. Notability is especially imortant because while NPOV encourages editors to add alternative points of view to an article, it does not claim that all views are equal. Although NPOV does not claim that some views are more truthful than others, it does acknowledge that some views are held by more people than others. Accurately representing a view therefore also means explaining who holds the view and whether it is a majority or minority view.

Soon it became evident that editors who rejected a majority view would often marshall sources to argue that a minority view was superior to a majority view - or would even add sources in order to promote the editor's own view. Therefore, the NOR policy was established in 2003 to address problematic uses of sources. The original motivation for NOR was to prevent editors from introducing fringe views in science, especially physics - or from excluding verifiable views that, in the judgement of editors, were wrong .

). In its earliest form the policy singled out edits that:

• introduce a theory of method of solution;
• introduce original ideas;
• define terms; or
• introduce neologisms

for exclusion, and established that

• ideas have been accepted for publication in a peer reviewed journal; or
• ideas have become newsworthy: they have been repeatedly and independently reported in newspapers or news stories (such as the cold fusion story).

as criteria for inclusion.

References

---

As before my attitude has been conservative, to try to preserve as much as the previous content as possible. I have attempted to make the layout more consistent and clearer. Slrubenstein | Talk 10:53, 13 September 2007 (UTC)

Let's see if it sticks, but first we need to unprotect the page. ≈ jossi ≈ (talk) 17:13, 21 September 2007 (UTC)

It may be a long time before we un-protect. I feel strongly we need to resolve the debate over primary/secondary sources, and that is likely going to take time. But like othes, I suspect, i would like to see forward movement. Perhaps if no one objects to this change, an dmin can make it without unprotecting the page. I am suggesting that if we can agree on a change, someone make it, while we continue to confront contentious issues. Slrubenstein | Talk 17:25, 21 September 2007 (UTC)

OK, let's follow process then, shall we? ≈ jossi ≈ (talk) 17:27, 21 September 2007 (UTC)

Vote for this addition to the policy page

Okay, this proposal has been sitting here for 9 days without any objections as to it's content. This proposal does nothing to change existing policy or any of the policy proposals. All it does is add a brief history of the policy and its evolution. I would like to move past this 'subject' so it can be archived, and this page can concentrate solely on more pressing issues. What is concensus here, so that I know whether or not I can ask an Admin to make this change? Thanks. wbfergus ^Talk 11:38, 22 September 2007 (UTC)

In case anybody misses the post near the bottom of this page, I have added the above "Origins" to the policy. If I was incorrect in there being a clear concensus on this, feel free to revert.

wbfergus ^Talk 13:20, 24 September 2007 (UTC)

Votes to ask for adding Origins of the policy

• Agree wbfergus ^Talk 11:38, 22 September 2007 (UTC)
• Agree provides the necessary context for an easier understanding of this policy, and it is brilliantly written. ≈ jossi ≈ (talk) 14:23, 22 September 2007 (UTC)
• Agree - time for some forward movement. WAS 4.250 15:02, 22 September 2007 (UTC)
• Agree - absolutely, clarifies the intent of everything else, gives meaningful context, and really not fault I can discern. SamBC(talk) 17:54, 22 September 2007 (UTC)
• agree - very clear and as Jossi points out it provides necessary context for users who have difficulty understanding the policy--Cailil ^talk 15:13, 23 September 2007 (UTC)
• agree well, obviously ... and reminding people that it has a narow purpose, to explain the origin of the policy in the context of the other two core content policies. Slrubenstein | Talk 16:59, 23 September 2007 (UTC)
• Agree This section is necessary, especially for new users or tendentious editors, in order to explain how and why we've gotten to this point. I made a few edits re typos and grammar, but made no edits that changed the meaning. In fact, other than a few things I would have phrased differently due to my own personal tastes, I think this section as proposed by Slrubenstein is excellent! &#0149;Jim62sch&#0149; 13:59, 24 September 2007 (UTC)
• Agree Understanding the background and rationale of the policy is crucial to understanding its importance. KillerChihuahua^?!? 20:45, 24 September 2007 (UTC)
• Agree I'm not sure if there's much for me to add, but this policy differentiates the original Wikipedia from POV-pushing copycats like Conservapedia and CreationWiki that do not rely upon this primary policy. This change improves the understandability, which is critical. OrangeMarlin ^{Talk• Contributions} 21:33, 25 September 2007 (UTC)
• Agree I second Orangmarlin's response, in that it does improve the understandablity on primary sourses. There have been so many instances, that I recall, where editors rely on these primary sources and add them to the article to support their POV, and then add another primary source when someone challlenges the source, therefore making it
  WP:SYN. This should be totally discouraged. I have to read more on this subject though, but do see a very dangerous pattern if this is not made clear on what a primary source is and what it means in order to sourse a claim made in the content of an article. Jeeny
  03:59, 27 September 2007 (UTC)

Votes against adding Origins of the policy

• Disagree. First, we shouldn't be voting on what the history of Wikipedia was. If this gets approved and I want to edit this, I should edit it and not have to face claims that the exact language is official policy we voted on. Second, nothing in this section is an actual rule, it's just commentary - an essay. It's reasonable to have a brief "rationale" section that explains what a policy is accomplishing today and how it fits in with other policies. But a policy page shouldn't have a secondary function of teaching history or context. Third, it's just too wordy and impertinent. Fine for an essay, too much digression and fluff for a policy page. Wikidemo 12:22, 22 September 2007 (UTC)

Many policies provide rationales for the policy. How the policy came into being is part of the rationale. The current policy has - and has had for a long time - an explanation of the origins. So having an origins section is not radical. My intention in this rewrite of the section was to be more informative and clearer. Could I have done a better job? Help out! As for editing it - well, for starts, if you have a sugestion right now, why not make it? This voting process is only (like most votes) a poll. There is no reason why it can't be combined with making suggestions to improve the proposed section. Slrubenstein | Talk 13:48, 22 September 2007 (UTC)

Careful what you ask for. Setting aside the spelling errors, long footnote, dig about an obscure competitor to Wikipedia, and extensive homage to Jimmy Wales (we certainly don't need any more of that in the policy pages), I would gut the historical context and say, if anything, why this policy is here and what it has to do with Wikipedia. I condensed this on the fly, and this would need more work, but I would replace the proposal in full with something like:

Reason for this policy

Wikipedia is a place where editors with diverse, often conflicting, points of view collaborate on the creation of an encyclopedia. As per

WP:NPOV

it is harder to agree on the truth than to agree on what people believe to be the truth. Therefore, Wikipedia does not address "truth" in any objective sense. Instead, Wikipedia is an unbiased compendium of notable schools of thought.

Verifiability

is the way Wikipedia ensures that the most significant views are represented, under the assumption that the most notable views were easiest to document with sources. Not all views are equally significant. Some are held by more people than others. Accurately representing a view requires explaining who holds the view and whether it is a majority or minority view. When an editor creates new material specifically for a Wikipedia article, and cannot cite whether any reliable sources share the same conclusion, we have no way to know if the view is significant or not, and therefor exclude it as original research.

(list of items is impertinent - if we're still banning them, include in policy)

-- Wikidemo 15:09, 22 September 2007 (UTC)

Nothing wrong to have, for a change, something in a policy page that makes perfect sense and that is wonderfully written. ≈ jossi ≈ (talk) 14:26, 22 September 2007 (UTC)

Comment - I think we should wait a while longer but if the discussion so far is any indication, there could well be a consensus to include this information. Can I please put in a request that if that happens we do so in a way that does not further complicate the organization and structure of the page? We also need to avoid redundancy. It's already fairly long and hard to read, particularly for some of the new editors we most want to reach. There are already multiple Jimbo quotes, footnotes, context statements, explanations, historical comments, etc. Could we do our best to consolidate these various things into sections labeled "history", "rationale", "background", etc., that are clearly set off from things like definitions and rules about what editors are supposed to do and refrain from doing? Thx.Wikidemo 14:35, 23 September 2007 (UTC)

I appreciate your constructive approach, and while I myself consider ways to edit down what I proposed, Let me just point out one thing: different people have different learning styles. This means there is a value to presenting the same information in different ways. Nothing stops a newbie from skipping a section that is unhelpful to them. For example, for people who want to get right to the point, we have the policy in a nutshell. Some people may think that is all we need to say, others would want far more explanation with lots of examples. I agree with you in general that the policy page should be well-organized and straightforward. Just understand that an explanation that you find confusing and irritating, others will find helpful and enjoyable to read - and vice versa. No one is wrong, people just have different learning styles. Slrubenstein | Talk 14:45, 23 September 2007 (UTC)

While I am generally given to a terse writing style, this is one instance where terseness would lead to more confusion. The genesis of the policy needs to be clearly and fully explained. (BTW: impertinent is decidely the wrong word.)&#0149;Jim62sch&#0149; 14:04, 24 September 2007 (UTC)

Comments

Good points Wikidemo. Thanks for those. After reading them I tend to agree, but I think the brief "history in a nutshell" may still be useful to the policy, if not actually part of the policy. Maybe either a sub-page or a separate article, so the concerns you raised are alleviated? wbfergus ^Talk 13:10, 22 September 2007 (UTC)

A little looking shows that

WP:NPOV has a (large) section devoted to the history of that policy however. I don't know if this will change anybody's opinion though. wbfergus ^Talk

14:45, 22 September 2007 (UTC)

Just to pick up on the protection issue. I think it is a little harsh assuming that the page has to be protected for the primary/secondary source debate now ensuing. This is not the same debate that there was edit warring over. I don't think anyone involved in the new debate has suggested radical surgery as yet, and even if they wished to, we could assert that it was a legitimate BRD manoeuvre. It would seem to me that the sensible approach would be to relax a bit: unprotect the page, apply the edit and see what unfolds. 'Tis the work of a moment to re-protect and revert (we could even agree that the rollback point is now if that would help). Spenny 15:37, 22 September 2007 (UTC)

I just added some hidden comment warnings to the policy, at the top of the page, and again at the top of each section, so that it is plain and unmistakable that there is a proper way and procedure before edits get made. Though it is redundant, the reason I added it in each section is so somebody can't claim they just hit the section "edit" button and didn't see the warning. Since these are hidden comments, I don't think it changes the policy at all, and refers to another policy about making edits to policy pages. The appropriate policy page/section is

, specifically where it states:

• A policy is similar to a guideline, only more official and less likely to have exceptions. As with guidelines, amendments should generally be discussed on their talk pages, but are sometimes forked out if large in scope. One should not generally edit policy without seeking consensus first. wbfergus ^Talk 12:26, 25 September 2007 (UTC)

I removed them, but I can explain... The fact that everyone can edit every page is a foundation principle; this applies to policy pages and articles alike. Certainly some edits should be discussed first, but not all edits need to be discussed before being implemented, and "this wasn't discussed" is a very weak edit summary to use when reverting an edit to any page. I have started to see that edit summary more often on some policy pages as of late, and this sort of warnings would only contribute to the mistaken perception behind that. — Carl (

CBM · talk

) 13:10, 25 September 2007 (UTC)

I will abide with what you and others think, but the bulk of my reasoning is on your talk page. I don't think I need to (or should) bother with copying it here as well. They weren't meant to prohibit any editing, just so any editors are fully aware of another policy about editing policy, but that's a judgement call I'll let more experienced folks make. Thanks for your reasoning behind it. wbfergus ^Talk 13:33, 25 September 2007 (UTC)

I think some concern about inappropriate edits is valid, but when truly inappropriate edits are made the right solution is to point that out on a user's talk page. An edit isn't inappropriate just because it wasn't discussed first, though, and I think there is no reason to add comments to justify people who revert changes only because the changes weren't discussed first. I left some longer comments on my talk page as well. — Carl (

CBM · talk

) 13:46, 25 September 2007 (UTC)

An edit that substantially alters the meaning, intention, or spirit of a policy page is not welcome unless first discussed and agreed in talk in a manner that demonstrates wide consensus of editors. If you make such edit, and get immediately reverted, that would mean that the change you made to established formulation of policy is not acceptable. If you make an edit and it is not challenged, that would mean that the edit you have made is incremental, does not change the spirit of the policy, etc. It is quite straight forward. ≈ jossi ≈ (talk) 14:09, 25 September 2007 (UTC)

<tongue_in_cheek>I suppose that when all else fails, one could always invoke the Wikipedia policy (which is still policy), "

Ignore All Rules

".</tongue_in_cheek>

Okay, seriously though, I've been playing catch up to where many of you admin-types have been for a while. It's interesting reading some of the old original policies, like [rules to consider] (September 17, 2001). From it, it becomes a little more clear on why some polices originated and what the original intent was, even though in some cases what exists now may seem to have very little in common with the original. wbfergus ^Talk 18:16, 25 September 2007 (UTC)

There's a tension between various policies, guidelines, and aphorisms about ignoring all rules, seeking consensus, being bold, being civil, not wikilawyering, etc. Perhaps a good thing. Rules like bridges work best when in tension. Regarding incrementalism, it's a good thing for policy and guideline pages to change slowly and that may be an end in itself, but taking too tough a stance on that can make you appear to be defending a walled garden (too nice a phrase to describe it - a fenced schoolyard?) and actually encourage rather than calm edit wars on the policy pages. Rather than stern edicts in the comments and when reverting, it's best to have the patience to point people to discussions on consensus, explain why, etc. Wikidemo 18:48, 25 September 2007 (UTC)

Dispute tag

I do not think that a few editors that disagree with long-standing formulation of this policy, can held this page at ransom by adding a dispute tag. This is official policy of Wikipedia, until that time in which there is consensus to the contrary. If editors want to alert others about their concerns, they can post a message at the VP. ≈ jossi ≈ (talk) 17:42, 25 September 2007 (UTC)

Actually, when there's a consensus to change it it won't be disputed any more, because it'll be changed... so what situation, exactly, is the disputed tag appropriate for? SamBC(talk) 17:46, 25 September 2007 (UTC)

When there is substantial support for a change, which is not the case. ≈ jossi ≈ (talk) 17:48, 25 September 2007 (UTC)

My advice is to ignore the dispute tag altogether. Whatever positive uses it may have once had it no longer enjoys because of chronic misuse. Just ignore it; the policy has no more or less weight with a "disputed" tag than it does without it. Editors who post a "disputed" tag will either gather consensus to change the policy, or they won't; in either case, in a few weeks, the disputed tag will no longer likely to be an issue. — Carl (

CBM · talk

) 17:53, 25 September 2007 (UTC)

This page is not official policy in itself. It documents official policy derived from Wikipedia practice and convention. If it were impossible to change a non-consensus section in a policy page without consensus, policy pages would never change, despite widespread practice to the contrary. Since there would always be a contingency devoted to the non-consensus description of policy, you'd be stuck with it forever. What matters is that the section is controversial, and there is ample proof of that. See Wikipedia:Policies and guidelines.COGDEN 18:09, 25 September 2007 (UTC)

Really, the choice is to either remove the non-consensus section, or to put up the tag. Or, the page can be protected back in the form it has been for the last several weeks. Editors have to be warned that the section does not reflect consensus, and might soon change. COGDEN 18:09, 25 September 2007 (UTC)

I would have said "may not" instead of "does not", since as far as I can tell, there has not even been an straw poll conducted to see how people feel about the issue. The major problem with asking for a straw poll would be trying to decide how to word it. A simple poll would surely fail. But a long-winded explanation of what is being asked and why and which alternative to choose, just as surley fail as well, if not properly worded, and containging viable alternatives so the the current NOR policy itself is not weakened. wbfergus ^Talk 18:24, 25 September 2007 (UTC)

Also, while straw polls can't be used to solve anything (per the page), it may at least give a general overview of how many people are on either side, for example "How many people think PSTS absolutely has to stay where it is unchanged" vs. "How many people will agree that PSTS will remain unchanged, but moved to another appropriate policy level page for further discussion", or something along those lines. Any talk, votes, or other moves to move PSTS to anything other than another policy-type page will surely fail at this point, as only a small minority can prevent concensus (from the way I read the rules anyway). I still haven't found anything yet that clearly states what concensus is in Wikipedia, only vague statements. Can anybody provide a link that clearly states what concensus is, like 2/3 of the opinions or anything like that? Otherwise, exactly how is concensus determined if things like straw polls aren't used to determine concensus? I'm not being a smart-ass, I'm really trying to figure this out. Thankswbfergus ^Talk 18:32, 25 September 2007 (UTC)

Actually, there was a straw poll several weeks ago

here. It's in the archives now. It showed that there is no consensus. COGDEN

23:00, 25 September 2007 (UTC)

The poll questions did not accurately reflect the statement of policy, and the bias of the questions was noted by several editors. (In the interests of disclosure [for the benefit of those not active during the poll], I was one of the editors who objected to the wording.) All it showed was there was a lack of consensus for severe restrictions that are not present in policy, and never have been (to my knowledge). Additionally, a number of editors have modified their positions due to discussion and a number of editors have joined the discussion that were not present then. Vassyana 06:53, 26 September 2007 (UTC)

There is no hard and fast rule for determining consensus in straw polls. What is more useful are the comments users make, which help shape consensus. That's why an RFC is better than a straw poll for helping resolve disagreements such as this. The current RFC, though, isn't formatted in a way that is likely to get a lot of participation. It would help to have an executive summary of the issue and a section devoted to RFC comments. — Carl (

CBM · talk

) 18:36, 25 September 2007 (UTC)

As a guess, I don't think there's a consensus for doing away with the source typing and I doubt those who oppose it will win enough people to their side for that. With some patience and advocacy, people might generate a consensus around changing the definitions and/or moving them somewhere else. If I'm right, is that enough for a dispute tag on a policy page (which seems unhealthy to me, at best a last resort)? How are we supposed to know if a dispute is legitimate or significant enough to report? Wikidemo 18:48, 25 September 2007 (UTC)

This is a long-established aspect of policy and has the consensus of many hundreds, if not many thousands, of editors and administrators who rely upon it in dealing with WP:NOR on a day to day basis. It will need a serious case to be made that more than a few questioners or disputants will be required to keep a "disputed" tag on any current aspect of WP:NOR. ... Kenosis 19:40, 25 September 2007 (UTC)

There are far more than a "few questioners or disputants". Moreover, the primary/secondary/tertiary model has never been true Wikipedia policy. The fact is, primary sources line peer-reviewed journals, interviews, works of literature, conference proceedings, movies, etc. are used far more frequently than secondary sources in actual Wikipedia articles. That's what the consensus is, and therefore that's the policy. When Wikipedia practice does not correspond to a policy page, it's the policy page that must change, not Wikipedia practice.

Moreover, you do not need a consensus to change a policy page that does not currently reflect consensus practice. If that were true, non-consensus statements in policy pages would be permanent no matter how controversial, since there would always be a constituency favoring the language. There is no such thing as a highly-controversial Wikipedia policy! COGDEN 23:12, 25 September 2007 (UTC)

I think you give too much credence to your own views. Absent of a substantial challenge (which the current one is not), the wording in policy pages is a reflection of current practice. ≈ jossi ≈ (talk) 23:18, 25 September 2007 (UTC)

This claim would be all the better for some evidence of present practice. The bald assertion that this page must magically reflect present practice seems unfounded; practice may well change while this page is still being reverted. Septentrionalis PMAnderson 23:38, 25 September 2007 (UTC)

Please stop adding and removing the tag. Such revert warring will not solve anything and will just make both sides less willing to discuss things. I have seen such arguments before... those who are defending the policy can say "prove to us that there is consensus to change"... those who want the tag can counter with "prove to us that there is consensus to keep"... Let's focus on what will solve this dispute instead of making it worse. Blueboar 19:46, 25 September 2007 (UTC)

The onus is on those wanting to change things to demonstrate there is a problem, given that that the current policy has been in place for several years, and there is consensus to change. Throwing tags around willy-nilly on policy pages by a small group of malcontents with some personal agendas does not constitute some serious dispute or serious problem. What is going on here?--Filll 19:56, 25 September 2007 (UTC)

Anyone following the talk page will know there is a problem. (Also read the last four or five archive pages.) No consensus is needed to change (or when all else fails, to remove) a highly-controversial section of a policy page. The fact that it's controversial means that it does not "describe common Wikipedia practice" (See Wikipedia:Policies and guidelines). If a consensus were required to change such sections, they would never change so long as some constituency wants it there, regardless of how controversial it is and how much damage it might do to the Wikipedia institution. COGDEN 23:18, 25 September 2007 (UTC)

What we really have is a few editors throwing a hissy fit that their proposed changes are getting nowhere and who are subsequently violating

WP:POINT by misusing the disputed tag. As a policy (not a freaking "definition" or whatever it was alleged to be of policy) it needs to be brought before the entire community before changes can be made, and cannot be held hostage to the whims of a few disagruntled and apparently misinformed Wiki editors. I don't see the community being in a lather about this policy: hence the tag was placed inappropriately, and was rightfully removed -- numerous freaking times. &#0149;Jim62sch&#0149;

21:10, 25 September 2007 (UTC)

The dispute tag marks the presence of a dispute - which is the absence of consensus - which is what we have here. The absence of a dispute tag should mark the absence of a dispute - which is the presence of consensus, either for the present version or for agreed-on changes. Removing the dispute tag falsely characterizes the section as being undisputed. Jacob Haller 20:20, 25 September 2007 (UTC)

Pardon me, but if I am not mistaken, are not policy pages a bit different than regular article pages? It strikes me that one should be less cavalier with dispute tags on policy pages. However, I might be totally mistaken here...--Filll 20:27, 25 September 2007 (UTC)

Indeed, the heading box at the top of the article claims that the items in the article reflects widespread Wikipedia practice. The absence of a tag turns the policy page into a lie, doesn't it? COGDEN 23:18, 25 September 2007 (UTC)

Tsk! Tsk! Go off to parents' evening, come home again to find all you naughty children have made this mess. I think you all should be sent to bed. It strikes me that however right or wrong the tag is, it is hardly worth warring over. Now say sorry and tidy your rooms.

Filll, please be careful about comments questioning peoples' motives. There has been an undercurrent of suggestion that those who want the policy explained in clearer terms are doing so out of malice which I thought was dissipating as people gained understanding of the issues; there are some people posting in superior terms about the qualities and qualifications required to participate here. For example, if I am right that some of this stuff is too complicated for your average editor either to be interested in or understand, then that group are unlikely to be participating in this current, if I may say, rather rarefied debate. But everyone deserves a slap on the wrist of common sense ;) Spenny 21:08, 25 September 2007 (UTC)

It's written at grade level 15; I would assume that anyone proposing to write an encyclopedia should be able to read at at least that level, especially given that many of our sources are written above that grade level. &#0149;Jim62sch&#0149; 21:17, 25 September 2007 (UTC)

No idea what grade level 15 means. I don't doubt that to write some parts of the encyclopaedia there is a requirement to be highly intelligent. However, Wikipaedia covers a very wide range of subjects, and until someone determines that the wide raft of other stuff other than the academic is unencyclopaedic, then I would suggest, and I have already suggested, that you need to ensure that the language is targeted at the simplest level possible - does policy not apply to Britney Spears? In the UK we have an organisation called the Campaign for Plain English which awards crystal marks for documents written in straight-forward English. I have suggested that it represents a goal Wikipedia should have, not just for policy, but for articles. Having an encyclopaedia written to a high grade of English is not something to be proud of: the art is to get the same point across in the simplest English. That takes real editing skill. Spenny 23:26, 25 September 2007 (UTC)

A bit of reasearch on your part (the type we often do for any article) might have gone a long way: see

Flesch-Kincaid#Flesch.E2.80.93Kincaid_Grade_Level

.

Also, if you read carefully, you will see that I noted that many of our sources are written above that grade level. In other words, no matter what level we may chose to be efficacious for the writing of a widely comprehensible encyclopedia, the sources we often use are written at a fairly high level and we need to be able to apprehend them in order to use them. Quite a simple concept really. &#0149;Jim62sch&#0149; 17:17, 26 September 2007 (UTC)

Well, let me join Filll in questioning the motives herein. My only problem with changing this policy is Evolution. The second the policy is gone (and it doesn't matter if it's the slightest change), POV warriors will be swarming in, and "upgrading" the article to Evolution is bad theory without any scientific support with references to some Creationist website. Holding this hard-fought ground is fine. So, there are no good reasons to change this policy. None. Yes, I've kept up with the discussion, as convoluted, confusing, and confounding as it is. OrangeMarlin ^{Talk• Contributions} 21:20, 25 September 2007 (UTC)

Nobody is suggesting changing this policy. Nobody is suggesting removing the policy. What we are saying is that the sound principles of the policy are at risk of being subverted by confusing explanations and improper tests of OR. So, there are no good reasons to change this policy. None. If you have been keeping up you will see that there has already been a consensus change where it was recognised that the OR policy had been subverted by inappropriate amendments before this current discussion started.

Is this policy page a perfect expression of policy, accessible to all? I suggest not. If not, then there are two good reasons to be looking to improve. Spenny 23:26, 25 September 2007 (UTC)

Hang on a second here... the slightest change, however well-thought-out, to this policy will immediately lead to the complete and utter entropy of the

WP:CIVIL. SamBC(talk

) 21:28, 25 September 2007 (UTC)

Probably a better waste of our valuable time to deal with

assume good faith, but it doesn't violate civility. Telling someone that they are "a #$%(*&ing POV-pushing @#$%head". Now that's uncivil! OrangeMarlin ^{Talk• Contributions}

21:38, 25 September 2007 (UTC)

Please refresh yourself on the civility page.

WP:CIVIL isn't rules, it is guidance. Spenny

23:41, 25 September 2007 (UTC)

Got distracted by your smoking. Back to Evolution. Without a strong policy and vigilant editors, the article gets messed up. We have to spend hours just finding the crap articles that people try to sneak into the project. So yeah, I am totally worried that POV-warriors will use any change in policy to lead a charge towards complete and utter entropy of the article. OrangeMarlin ^{Talk• Contributions} 21:40, 25 September 2007 (UTC)

I share your concerns about the

WP:FRINGE, which this policy page doesn't address. Essentially everything that can be said against evolution by creationists has already been said before, numerous times. Indeed, the more "secondary" such sources are, the wackier they become. It's never original. And it never depends on whether your sources are "primary, secondary, or tertiary". If it's unverifiable fringe research, it doesn't matter whether you get the fringe research directly from the wacko researcher, or you get it second-hand from some evangelical commentator. COGDEN

23:26, 25 September 2007 (UTC)

If there is a dispute, there is very likely to be an absence of consensus. Since {{policy}} loudly asserts wide consensus, it is very useful to label the points where this is an exaggeration. As it happens, I think we need to have source typing in one of core policies (probably this one, since writing from primary sources is one quite common type of OR); nevertheless, I support the tag. Septentrionalis PMAnderson 23:01, 25 September 2007 (UTC)

Brief aside: using primary sources doesn't make it automatically OR. I think the recent discussions here have pretty well established that. SamBC(talk) 23:03, 25 September 2007 (UTC)

Nor do I say it does; but use of primary sources only, as in the embarassment recently on the front page,

WP:SYNTH. In that case, the article managed to omit that its most quoted source is a piece of poetry, written to flatter, and in a genre which did not value historic accuracy. It may be that the primary author understood the problems involved, but he failed to make them clear to the reader. Septentrionalis PMAnderson

23:09, 25 September 2007 (UTC)

Even using primary sources only is not in itself OR. Many, many Wikipedia article rely only on primary sources such as interviews, novels, movies, works of art, peer-reviewed journals, conference proceedings, speeches, comments by press agents, even poetry. Your concern about the

NPOV. The poetry, and non-critical acceptance of the poetry, isn't "new research". (It's very old research.) But in this case, statements about the poetry, and implications made by editors, do not include the critical POV that the poem is fanciful. COGDEN

23:37, 25 September 2007 (UTC)

One revert rule

I don't know if anyone appreciates the irony of the edit war over the disputed tag at the top of the PSTS section resulting in a protect tag at the top of the page saying there is a dispute. It seems to me that this is a worse outcome, sort of a self-fulfilling (or self-defeating) prophesy. Anyway, I suggest we agree in advance that when the protect tag expires, we all follow the

00:19, 26 September 2007 (UTC)

There's another related irony, which is that we're having a fierce edit war by numerous editors over whether we should admit that a section is controversial. I participated in the last edit war (I think I reverted twice), but this time I only reverted once, because I could predict what was going to happen: Between each edit war-protection cycle, more people join the controversy on each side who couldn't possibly have kept up with the reams of archived arguments written on this issue, we re-hash the same arguments, and any good movements toward consensus, like that managed most recently by Blueboar, get forgotten. Then, by the time the page comes up for automatic unprotection, we're back to square one. I think the first step toward recovery is to admit we have a problem. There's fierce disagreement on the primary-secondary source model. And the sooner we admit that, the sooner we can start healing. COGDEN 00:46, 26 September 2007 (UTC)

If people don't realize the sheer irony of edit warring over a disputed tag, 1RR is essentially meaningless. Also, there's quite enough people involved here on both sides to make 1 revert per person worthless. How about we all agree to not edit war over a policy or be banned from this policy page for three months? That would actually have teeth, provided some admins have balls enough to enforce it. I'm a bit cynical about the dispute and some people involved at the moment, so take those comments with some grains of salt. Vassyana 02:00, 26 September 2007 (UTC)

That's a good idea, but I think we need some sort of consensus language in that section first. The current section is broken (it's controversial, which by definition makes it wrong), and nobody should be sanctioned for trying to fix it in good faith. But when we arrive at some rough consensus language, further editing at that point ought not to happen without prior discussion on the talk page. COGDEN 17:33, 26 September 2007 (UTC)

Can we agree to at least maybe have something along the lines of a [[straw poll to determine which cources of action would be acceptable for further discussion, so that we aren't spending our limited "editing time" arguing about some point that may become readily apparent doesn't have a snowballs chance? Something along the lines of Wikipedia:Editing policy pages maybe, where 8 different 'subjects' or proposals are put to a non-binding vote. It should be clearly stated that the votes won't be used to determine yea or nay to the policy, just to try and determine if there is some common ground for further discussion. Good or bad idea? wbfergus ^Talk 18:55, 25 September 2007 (UTC)

I don't know... perhaps we should conduct a poll to determine if we should conduct a poll about whether having a poll is a good idea or not. Seriously, having sat through the "poll" at ATT, I have a feeling that it will just add to the confusion. It is obvious that there are more than just "a few" people who have problems with the wording (if not the concept) of the PSTS section. The question is if there are enough people who feel something is wrong to warrent a "disputed" tag. I don't there is a "lack" of consensus for the policy... But, I do think there is a problem that needs to be addressed. Blueboar 19:40, 25 September 2007 (UTC)

Before we go around and let 3 or 4 people with personal agendas change the principles under which WP has operated for years, lets try to understand what the issues are. I think there is no consensus to change, and this is quite clear. What on earth is the problem with the current status quo?--Filll 19:54, 25 September 2007 (UTC)

Try going back and reading the comments on this page and the last two archive pages, Filll... Then go back and read the comments of several dozen people who have raised similar issues for more than a YEAR. There may not be a consensus to change, but there sure as hell are more than just "3 or 4 people with personal agendas" commenting that they have concerns about this section. I assure you that I do not have any agenda... I am a firm supporter of NOR. But I do have concerns about the current wording of PSTS, and how it ties in with NOR. Blueboar 20:07, 25 September 2007 (UTC)

Filll, before you wade waist-deep into a debate that has finally gotten reasonably friendly again, please try to familiarise yourself with the discussion that has already taken place. So far as I know, there is no-one with a "personal agenda" to change the policy. In fact, all that's being debated is the expression of the policy. No-one has advocated removing or substantially weakening NOR here for quite some time, although some people believe that certain good-faith suggestions would have the unintended effect of doing so. Most of those issues have been reasonably well worked through now, in fact, and throwing around accusations is not going to help the situation. SamBC(talk) 20:13, 25 September 2007 (UTC)

Frankly, the fact that I am advised to dig through months if not years of history and the dispute cannot be stated succinctly and clearly, tells me something smells bad here. So you are telling me all this hubub and sturm und drang here is because a few mean old editors want to stay with the old policy, when for years people have been clamoring to change it? Hmm...--Filll 20:25, 25 September 2007 (UTC)

No, and it's misleading to characterise it as one or the other. Sensible, reasonable people are taking various positions, none of which represent an intention to sabotage policy. There was a reasonably summary not long ago, in a section entitled something like "for people just arriving". It's a little out of date now. However, expecting a new summary every time another editor arrives to join in the discussion is a sure way to stop discussion from actually happening. SamBC(talk) 20:30, 25 September 2007 (UTC)

What happened to that essay setting out the case for change? .. dave souza, talk 20:52, 25 September 2007 (UTC)

Consensus? It's more like 2-3 editors want some change for some unknown and undecipherable reason, and want us to agree? Uh, but pardon me if I don't agree.OrangeMarlin ^{Talk• Contributions} 21:14, 25 September 2007 (UTC)

More than 3 editors have advocated the idea that a change is necessary, or would be beneficial. A number of editors who were initially resistant to the idea have concended some merit, and once again some editors who haven't participated in the conversation make comments that bear a striking resemblance to

WP:AGF shown in impugning the motives of the editors raising issues. For crying out loud, what motive do you actually think any of us might have? SamBC(talk

) 21:25, 25 September 2007 (UTC)

<undent>I went back and all I see is a confusing mess. I am not going to spend 10 hours digging through obfuscation and nonsense and contentious disputation to try to find the nugget of the argument, when people do not seem to be sufficiently motivated to state it here where we can read it. This looks suspiciously like you want to drive others away so you can do your dirty work in secret. Either make it clear, or forget it.--Filll 21:39, 25 September 2007 (UTC)

One can only state clearly that which is ratiocinated clearly. &#0149;Jim62sch&#0149; 21:57, 25 September 2007 (UTC)

OK, let me take a stab at it, for what it's worth. An edit war developed over a statement that favored secondary sources over primary. The edit war lead to several cycles of protection, during which the discussion examined first the statement, then the definitions of primary and secondary sources, then the need for source typing in the context of NOR. One of the things that came from the discussion is that it is not the type of source that creates NOR problems, it's their use. If you look above, you will find different drafts explaining the use of sources in relation to NOR without classifying them as primary and secondary sources. The discussion also examined whether the definitions of PSTS belong in NOR, and various alternatives were discussed including moving them to another policy page, a guideline, or an essay page. Then new editors came to the discussion here to oppose any change. And here we are. Dhaluza 01:34, 26 September 2007 (UTC)

Filll and Orangemarlin, I really (honestly and truly) expected far better from the both you than snarky comments mischaracterizing the situation, assuming bad faith and disparaging more than a few intelligent and experienced editors. I'm seriously and frankly shocked. Vassyana 01:42, 26 September 2007 (UTC)

One problem with OrangeMarlin's Filll's and Jim62sch's holier-than-thou comments is are that this very subject we have debating here recently has been problematic in this policy since at least January 2005 (see further instructions below, I can't get it linked) the oldest talk page I can find. Numerous people have argued these points over and over, and it appears that they probably just got tired of being stonewalled by groups oppesoed to any change not of their own doing, and figured that their time could be more constructively used elsewhere and they would edit as they saw fit, regardless of what the exact wording of the policy actually was. If their edits got reverted or challenged later, oh well, there was still a good chance most of their edits to artciles would still stand. So, basically what I'm saying is it seems to kind of difficult to effectively articulate 30 archived talk pages (some are buried underneath the draft rewrites), into a one or two sentence position, as he would suggest. Conversely, I've yet to see any effective or coherent reasons to leave a problematic section that has generated much controversy and heated discussions over almost three years (just about 3 months shy). wbfergus ^Talk 10:43, 26 September 2007 (UTC)

To see the page, go to the Archives, and select the very first one, "No original research (draft rewrite 5th December 2004 to 5th February 2005)", then select the "Discussion" for that page. wbfergus ^Talk 11:22, 26 September 2007 (UTC)

Shouldn't have struck though "is", problem is the subject. In any case, if these points have been argued, it's quite obvious that the arguments have gained no traction simply because they are not seen as valid arguments by the community as a whole. You seem not to realise that ultimately it is the entire Wikipedia community that decides these issues, and while there may have been complaints, they were quite likely soundly rejected given that the policy didn't change (I'm guessing here since you've provided no diffs and I'll be damned if I'm going to read through a whole damned archive section).

Holier-than-thou is both funny and out of context. &#0149;Jim62sch&#0149; 17:25, 26 September 2007 (UTC)

This policy would be more clear if it said what needs to be said without using the words "primary" and "secondary". - WAS 4.250 22:12, 25 September 2007 (UTC)

I may agree with you, and invite you to start a sandbox page in which these ideas can be explored. ≈ jossi ≈ (talk) 22:55, 25 September 2007 (UTC)

We already have some good proposals that describe the policy and practice without using the primary/secondary/tertiary concepts. I think we should start there. Or go back to them. With such a high volume of comments, it's easy to lose track of the arguments and the proposals. COGDEN 00:53, 26 September 2007 (UTC)

Agreed. I think this is the only way toward consensus. We all agree (I think) on what the policy is. What we argue about is how to express it, and about philosophical side-issues like the meaning of "primary" and "secondary". I think we're trying to be too cute by using such high-falutin historiographic ideas as primary and secondary sources. COGDEN 00:53, 26 September 2007 (UTC)

Make a proposal in a sandbox page. ≈ jossi ≈ (talk) 00:56, 26 September 2007 (UTC)

Jossi, I don't know what good it will do, but I am creating

Wikipedia talk:No original research/1 as a sandbox page and will put the examples in it that were made on this page as examples of explaining NOR without using the words "primary" and "secondary". Please note that these examples were not written with the intent of being perfect replacements for the section in question, merely proofs of concept that such a thing makes sense. Further it seems consensus that the primary/secondary/tertiary distinction is useful to keep somewhere in Wikipedia policy space even if NOR is better off explained without it. So please, no one moan over that straw man. WAS 4.250

02:35, 26 September 2007 (UTC)

I'm sorry, is the argument here that "primary" and "secondary" are too hard or too complex to be understood? Obviously "tertiary" must be a real poser then. Good thing we don't have quarternary sources: just imagine the mental agita they might cause. &#0149;Jim62sch&#0149; 16:58, 26 September 2007 (UTC)

<undent> I agree with Jim here. Holy cow! If this is a problem, then explain primary secondary and tertiary with some well-chosen examples. This is an unbelievable objection...wow.--Filll 17:04, 26 September 2007 (UTC)

I don't know if anybody is seriously promoting the idea that the concept of tertiary sources should be part of the OR policy anymore. Maybe I'm wrong: is that still a sticking point with anybody? COGDEN 17:38, 26 September 2007 (UTC)

Missing the point. Anyway, the T in PSTS stands for tertiary. Are they propsed for elimination? &#0149;Jim62sch&#0149; 20:11, 26 September 2007 (UTC)

I was just saying, not trying to respond to the main point. I was just thinking that I hadn't heard much on the tertiary issue for a long time, and it seems like maybe we at least have a consensus that the tertiary aspect of the historiographic OR model is not helpful. We'd call the remaining model the "PSS" model. Maybe this is one small step in the right direction toward consensus. COGDEN 21:50, 26 September 2007 (UTC)