Wikipedia talk:No original research/Archive 25

Archive 20Archive 23Archive 24Archive 25Archive 26Archive 27Archive 30

Clear and simple (2)

The text in question has been part of this policy for many years and it will not be removed unless there is wide consensus for its removal. I count maybe five or six editors that want to change this text and that are arguing for a dispute tag to be added. I say, no, you cannot do that. If these editors want to make a proposal for change, create a sandbox page as a proposal, and advertise profusely in VPP, the mailing list and all talk pages of core policies, so that the community is aware of your proposal. Then sit back and listen to its voice. That is the way to address significant changes in policy pages, such as removal of entire sections, changing the underlying principles of policies and other such changes. ≈ jossi ≈ (talk) 22:24, 25 September 2007 (UTC)

Except the discussion has moved on from removing that section, to refactoring and possibly relocating it. The dispute tag, as I understand it, refers to the wording, not the existence, of the section. SamBC(talk) 22:44, 25 September 2007 (UTC)
For those who want a summary of the reasoning, here's a very brief version. There have been a number of problems on article talk pages (specifics have been brought up here, and I'm not going to dig out the specifics right now) with people exhibiting confusion about the implications of the PSTS section on discerning original research. This lead a number of people to seek to clarify the presentation of it. After much debate with very differing views around, it seemed that presenting the same principles without recourse to the ideas of primary, secondary, and tertiary sources was sensible. With widened participation after such was sought on the village pump, cogent arguments against the loss of the section were presented. Since then, discussion has (mostly) focussed on refactoring or relocating the section, along with the addition of material that was originally conceived to replace it, but has been argued to be of benefit to the clarity of the policy in any case.
Some people are still arguing that it would be best to lost the section completely, but this no longer has any real degree of support. SamBC(talk) 22:44, 25 September 2007 (UTC)
I've been looking for an affirmative defense of WP:PSTS, but I have not seen one yet. I may have missed it in all the back and forth, but I don't think I have seen the cogent arguments you refer to. I have seen people saying the section developed over X amount of time, and has been around for Y amount of time, but of course it was not around for Z amount of time, so all this may be of historical interest, but of little practical value. I've also seen various assertions of its importance, but no serious discussion of the costs vs. benefits. I've also seen assertions that PSTS is supported by a really large groups of editors, and have seen the questions and concerns about PSTS dismissed as the work of a really small group. None of this rises to what I would consider the level of cogent argument. It seems to me that a really large group of supporters would produce a really large pool of cogent arguments. Dhaluza 01:04, 26 September 2007 (UTC)

It happens that I think the section should (1) go on another page (2) include more clarification regarding different uses of the terms "primary source" and "secondary source" and, if Wikipedia-wide standards are needed, specificly state where the Wikipedia-wide standard differs from academic standards in different fields and (3) include more information on appropriate and inappropriate uses of sources with regard to NOR, probably NPOV, and possibly other policies. It happens that I also think the tag is appropriate to note the ongoing debate. I suspect that if I did not seek transfer and revision, I would still seek appropriate tagging. The section is disputed. Requiring consensus (as opposed to dispute) to add dispute tags would make the dispute tags useless. It would mean we could only mark disputes where there are not disputes. An awful lot of people have passed though this debate, on all sides - some people on two or more sides. Jacob Haller 23:36, 25 September 2007 (UTC)

Contra Jossi, the text has not been here for "several years." Another editor linked to several versions of PSTS to show its development in time. The passages which triggered the latest stage of the debate - the passages which some editors were using to remove primary sources and statements referencing to primary sources - were inserted some months, not years, ago. The debate expanded to include the appropriate uses of primary sources, the ambiguity rooted in different definitions of primary sources, etc. Jacob Haller 23:36, 25 September 2007 (UTC)

Make a proposal in a sandbox page. ≈ jossi ≈ (talk) 00:56, 26 September 2007 (UTC)
I'll simply endorse Blueboar's proposal - both halves. I might consider expanding PSTS with discussion of different understandings of "primary" and secondary" and new sections of appropriate use of each class of source, if PSTS is moved off NOR, but I'd keep things as short as possible, if PSTS stays on NOR. Jacob Haller 02:50, 26 September 2007 (UTC)

Comment. The text has not been hear for years. It's been here since October. Changes leading up to that point were made in the months preceding. You know I agree, and have vigorously argued, that the section did demonstrate consensus (at least in the past) and was the result of steadily moving in that direction. I've also made a few proposals and a ream of comments. I just feel lost and hardly even know what to say anymore at this point besides "fuck it, whatever happens to that damn policy, happens". Vassyana 01:39, 26 September 2007 (UTC)

Some of your proposals were excellent. Why don;t you bring them up here for discussion? ≈ jossi ≈ (talk) 02:12, 26 September 2007 (UTC)
I will do so when I get my head on straight in regards to this discussion. Thanks for the vote of confidence on my ideas. :-) I'm just seriously flustered with the behaviour of some editors and at the endless circles we're moving in here. I'll review my proposals for the most appropriate ones and post when I find my center, and therefore some good faith in the process again. Vassyana 04:36, 26 September 2007 (UTC)
I think it's understandable, given the events of recent hours, that some of us are feeling rather burned. I think this is especially true for Vassyana, who's put in a lot of work and articulated several sides of the debate over recent weeks, showing an uncommon ability to work with others towards a sensible solution. It's very demoralising when it seems that, whatever happens, every so often more new hands will show up to the debate and insult us and question our motivation. I reckon that Vassyana's invested more in this than I have, and I know that I feel burned by it. It begins to seem likely that the idea of near-infinite inertia of policy will win out simply by the emotional battering that editors get for suggesting that the policy page could be improved. SamBC(talk) 02:19, 26 September 2007 (UTC)
You could help by avoiding the use of "us" and "them" in your arguments. All I am saying is: put forth proposals that can be assessed on their merits. ≈ jossi ≈ (talk) 03:42, 26 September 2007 (UTC)
In the case of "them" and similar, I was attempting to avoid names, as I don't want to get into specific accusations. In this case, "us" is those who have been unfairly accused and impugned, and "they" are those who made the accusations. As to your actual suggestion, I completely agree. My statements were to try to give context and explanation to the strained tone that has become evident from several of "us" (in the same sense as I mentioned). I don't suggest that any of the unfortunate statements that have led to a feeling of being burned came from you, Jossi. SamBC(talk) 04:02, 26 September 2007 (UTC)
Jossi, when people constantly impugn one's good intentions and motives, a siege mentality (us vs. them, or "they are out to get me") is a natural result. I think it's insulting to everyone involved. It's especially rude to people like myself and a few others here who started off being extremely supportive of the PSTS distinctions and changed opinions/positions based on discussion and/or compromise. As you well know, since you helped me, I even advocated a tightening of the section not too long ago. I find it rude and ignorant, to be pretty polite, to have aspersions cast on me for believing that the PSTS distinctions cause confusion and supporting reform. Vassyana 04:36, 26 September 2007 (UTC)
It was just mentioned again that WP:PSTS has only been in existence since October 2006. Is the proposal to go back to the roughly three-and-a-half-year-old "Primary and secondary sources"? If, for sake of argument, the proposal is to move WP:PSTS off to another page, is the suggestion that it should instead be transcluded onto this policy page? Or is the argument that it should be moved entirely to another page? Other than a determination that the policy makes no sense to a number of people, and/or that some articles are having problems with it, what's the present suggestion here? ... Kenosis 05:56, 26 September 2007 (UTC)
Both transclusion and seperation-and-wikilinking have been suggested. I would suggest seperation with judicious use of noinclude tags, so that the key points can be transcluded whilst creating a new page that can go into greater detail with more examples and so on, with the full version wikilinked. SamBC(talk) 09:35, 26 September 2007 (UTC)
I have no objection to this approach, so long as the current content under PSTS is continued until that page gets a foothold and so long as no major changes are made to the basic concept. Expansions on a separate page would be helpful, IMO, to further issues such as how introductory texts can be tertiary sources, and a number of other things that I personally would not want to see shoehorned into that section. ... Kenosis 12:34, 26 September 2007 (UTC)
Much of the primary-secondary-tertiary model has already been moved to Wikipedia:Classification of sources. We can always link to there from this page, but this page has to be a true policy page in its own right. Although it can link to a non-consensus essay, it can't have any non-consensus sections. That's why we need to work out some consensus language here as a replacement. Then, if we want, we can link to [[Wikipedia:Classification of sources, which might then evolve into a guideline, assuming it accurately describes current Wikipedia practices. COGDEN 17:46, 26 September 2007 (UTC)
The page that says "::This document is extremely rough and not of any use for anything. Don't link here except from talk pages."? .... dave souza, talk 18:01, 26 September 2007 (UTC)
Indeed, the structure and purpose of that page is just about right, but the content is a bit "out there" and, aside from that, needs a deal of polishing. It might be worth replacing the PSTS section of that with a verbatim copy of the PSTS defintions from NOR and then working on them, if the page is to be put to the purpose now being discussed. SamBC(talk) 18:06, 26 September 2007 (UTC)

Whatever happened to Blueboar's proposal?

Way back, Blueboar proposed to move PSTS to another page and add Citing Appropriate Source Materials (alternative suggested section title: Sticking to the Sources) to NOR.

Are there any objections or improvements to the second part of Blueboar's proposal? Jacob Haller 02:08, 26 September 2007 (UTC)

In my opinion, it's pretty much a non-starter, not because it's wrong, because it isn't wrong. But it doesn't work as presently proposed w.r.t. WP:NOR. "Citing appropriate source materials" is for WP:V, if anywhere. Here in WP:NOR, the WP:PSTS distinction is a framework inextricably connected to how WP:NOR is applied. The words "stick to the sources" were just added to the text of WP:NOR, and there's no reason to think they don't effectively convey the concept of WP:NOR to readers of the policy page, at least not thus far. The concept of sticking to the sources, by itself, does not lend to a complete explanation of WP:NOR where there are conflicting views among article editors about how to express what's in the primary source(s), i.e., in the published source(s) closest to the origin of the topic . This is especially true 1) where many differing primary sources on a particular topic of discussion are in wide disagreement, or 2) where the editors cannot agree on the plain meaning of the primary source(s), and/or 3) where there are significant analytic, synthetic, interpretive, explanatory, or evaluative claims involved in the point(s) presented in the article. In these cases, secondary or tertiary sources must be used in lieu of primary sources. WP:PSTS (or at minimum something like an updated version of the older "primary and secondary sources" with, e.g. "derivative secondary sources") is needed in WP:NOR, not somewhere off on another page with a link. In other words, WP:PSTS not supplementary, but integral to WP:NOR. Wikipedians have come to be accustomed to using it and t's been successfully applied in every type of topic on the wiki, irrespective of that people occasionally have problems with it. ... Kenosis 06:32, 26 September 2007 (UTC)

Kenosis, it's clear that you find WP:PSTS to be a useful tool, and some Wikipedians agree. But myself and others disagree that is integral, or that the success outweighs the problems as you assert. I think your conclusion that it has been successfully applied to every type of topic is incorrect. It has already been pointed out that scientists and historians use the primary/secondary distinctions differently, and this has caused problems in interpretation. Blueboar's proposal, whether you agree with the exact language or not, shows that PSTS is not integral to explaining NOR by explaining NOR in relation tot the use, not the nature of sources. The problems in the application of PSTS for source typing have been demonstrated time and again on this talk page when editors in good faith could not agree on the precise definitions of primary/secondary and how to apply them to specific cases. The result is that the discussion moves from NOR to discussions of PSTS, and neither is resolved. PSTS is a distraction. If editors who have taken the time to study and discuss PSTS cannot reasonably agree on what it means, how are editors at large to successfully apply it on the wiki? Dhaluza 09:48, 26 September 2007 (UTC)
I'm not sure how "occaisional" the problems with it are, but they have been considerable even if they are rare. This indicates that something needs to be fixed. I did support removal (and relocation) of PSTS, although I would now personally advocate a strong refocussing of the section. Also, if people have become accustomed to using something to discern OR that isn't actually very good for that purpose, that's a problem in itself. Recent discussions still seem to indicate quiet disagreement as to whether primary sources are actually acceptable in any but the rarest circumstances. SamBC(talk) 09:38, 26 September 2007 (UTC)
I agree with Kenosis. It is often an important point of reference in identifying and explaining original research. Slrubenstein | Talk 10:22, 26 September 2007 (UTC)
Slrubenstein, this is the crux of the whole thing. Please give the best example (make one up) you can of a situation where the word "primary" is an asset rather than a liability in communicating the reason that you reverted someone's edit. WAS 4.250 11:28, 26 September 2007 (UTC)
I've only rarely seen it on an edit summary; mostly on talk pages. Given a few days of spare time, I'm fairly confident I could list articles in every, or virtually every, topic area on the wiki in which the editors have succussfully used it to sort out content issues in their articles. And given adequate time, I'm sure I could show a vastly longer list where it's the actual practice. Ordinarily, we simply call it "OR" or refer to WP:NOR. ... Kenosis 12:09, 26 September 2007 (UTC) ... And in response to comments a bit farther above, people have problems with VER and NPOV too. That's not good cause to change those core policies in any substantial way, because they work well on the whole right across the entire wiki. And w.r.t. this policy, even a local consensus here on the talk page would be quite inadequate to implement this degree of a change to longstanding WP policy. ... Kenosis 12:23, 26 September 2007 (UTC)
Nobody is advocating a change in policy. The policy is what it is, and I dare say that given an example, we would all apply the policy the same. The policy is not the policy page: it's current accepted Wikipedia practice and convention. What's at issue here is the current expression of that policy on this policy page. That expression does not have consensus, and therefore is not binding. We need to make it binding. We need the legitimizing stamp of consensus here.
When we arrive at consensus language, nothing will change in actual Wikipedia practice. We'll still all be citing interviews, quoting novels and movies, citing peer-reviewed publications and conference proceedings. We'll be following NPOV. Fringe ideas will be excluded. Nobody will be including original research on pages. The only difference will be that we won't be arguing over which sources are "primary" and which sources are "secondary". We'll be arguing over what is, and what isn't, "original research". We'll arrive at the same results, but with less arguing, and with less misunderstanding and metaphysical hand-wringing. COGDEN 17:56, 26 September 2007 (UTC)

Derogatory statements, and personal insults

Since this 'area' has been cropping up quite often in the past few days from people who haven't participated in these discussions at all within the last month, let me state for the record that it is extremely disappointing to have a few people with Admin status come into the middle of the discussion, make absolutely no attempt to determine what the discussion is about, and then cast aspersions on those who are advocating change, without even knowing what the change entails. A couple others have merely popped in long enough to simply say "I oppose" or "I disagree", without any further comment or suggestions on what may change their mind or even what their concerns are.

This extremely piss-poor behaviour on the part of a few over the last couple of days has led to (once again) a siege mentality of us vs. them, even after several people originally opposed to change had changed their minds after a long debate and things had started to get civil around here again.

So, once again, even though some progress was finally getting made, albiet slowly, we are once again getting divided into three different camps, as Blueboar elaborated on above, They are (slightly modified):

  • 1) Those who have problems with the language of PSTS. (Change the language group).
  • 2) Those who don't have a problem with the language of PSTS per se, but do not think it fits in the WP:NOR policy. (Just move PSTS to a better 'home' than WP:NOR).
  • 3) Those who feel that PSTS is vital to enforcement of WP:NOR as they can't explain NOR (or even OR) without falling back on PSTS, and do not think it should be touched in any way shape or form. (I oppose any and all change group).

We were finally getting groups 1 & 2 two to collaborate and work on ways to alleviate the concerns of group 3. Then, new members of group 3 come into the discussion and are immediately opposed to any change, without even knowing what the change is. They come in and start throwing around insults and refuse to make any constructive comments. Now, once again the policy itself has gone into protection because of edits wars over something so stupid and assinine as to whether or not the policy should have a tag saying a section is in dispute, when the section has been in dispute one way or another (language or substance) since at least January 2005 (as can be seen in the archives). wbfergus Talk 12:52, 26 September 2007 (UTC)

RE the last sentence of the first paragraph of wbfergus' comment above ("A couple others have merely popped in long enough to simply say "I oppose" or "I disagree"..." ): Presumably they are doing other things, either within WP or elsewhere, and consider this policy already in place w.r.t. the work within WP. I cannot effectively speak to the reasons that some of the "critics of the critics", so to speak, may have chosen not to provide additional substantive commentary. Sometimes the word "no" is enough. For anything else about the recent arguments here, I think one would need to ask them individually by mentioning usernames and providing specifics. Regarding the last paragraph just above, agreed that some snipes and exaggerations appear to have gone, in more than one direction.

W.r.t. the 1,2,3, bulleted grouping, it neglects the very important option of moving WP:PSTS to another page for possible expansion of the concept while transcluding the present content back onto this page, allowing any necessary additions, clarifications, and/or minor modifications to be made to PSTS. I believe (correct me if I'm wrong) that Dave Souza proposed this particular approach. I also believe several others, myself included, have indicated they thought that moving WP:PSTS to another page with transclusion or partial transclusion is a reasonable course of action. ... Kenosis 13:19, 26 September 2007 (UTC)

Thanks Kenosis. Regarding the possibility of transclusion or other similar type discussions, I think that is where the bulk of the people in the first two groups have been working, on trying to find some acceptable way that this can be accomplished so people like you who say there a demonstrated need for the PSTS section at a policy level is the only acceptable alternative, though you seem to acceptable to the idea of allowing the move even to another (or new) policy if proper linking or other inclusions are in place so that it is still easy to determine that while not an actual part of the policy, in essence it is still enforceable by the policy. I think I said that right, if not, please correct me. wbfergus Talk 13:40, 26 September 2007 (UTC)
No, it is an integral part of the policy. Removing it from the policy is a non-starter, lacking a coherent replacement where most everyone who sees it agrees it's superior and more accurately reflects how WP:NOR is implemented across the wiki. ... Kenosis 15:43, 26 September 2007 (UTC)
I'm glad to see that you are open to replacing the primary-secondary model with superior equivalent language. I think that's progress. Maybe we can work together to find out what that language us. As far as I can tell, the closest we've come to consensus language is the proposal managed by Blueboar. What's your stand on that proposal, and how would you improve it? COGDEN 18:32, 26 September 2007 (UTC)
I already gave my opinion just above and in WT:NOR#Whatever happened to Blueboar's proposal?. IMO, we've yet to see anything close to a reasonable proposal for a replacement for WP:PSTS. Even if most or all participants here could agree that it was a significantly improved approach to PSTS, which we have not, there would still be a great deal of work to do to persuade all those users out there that've been using WP:PSTS in their discussions and practice. I also gave my opinion about Bluebear's proposal in particular. Most of it doesn't belong in WP:NOR, but in WP:VER, if anywhere. I inserted the words "stick to the sources", from Blueboar's proposal into WP:PSTS. Beyond that, I wouldn't know what to recommend changing in it so that it helps WP:NOR. Like i said, if someone comes up with an idea that looks to me like it will help residual confusion about WP:NOR without trashing the PSTS or PSS distinction, I'll be more than pleased to help. Thus far, IMO, that hasn't happened, and it appears I'm not at all alone in that opinion. .... Kenosis 22:33, 26 September 2007 (UTC)

As I am one of the editors who supposedly "popped in here" and who is accused of insulting the editors here who are diligently working away, who did I insult? If I stepped on your toes, I apologize. I wanted to understand what makes people so frantic to change a policy that seems to be working fine. I have suspicions of ulterior motives. I will be looking for examples of how the current policy is failing, and some new policy would improve things. I was met with a fair bit of rudeness. I guess my comments "shocked" some people. You should not be shocked if you understand me, and my editing philosophy and my concerns for some of the agendas and motivations here. In fact, I think it should be blatantly obvious, but I am not going to make that explicit if others cannot figure it out. In any case, people did state what they think the problem is, and hopefully I will be able to see some examples. I agree with the people above that instead of developing a small knot of people here whining, they have to make this case far more explicit and detailed, and put material in a sandbox for inspection, and then, if it comes to that, advertise for weeks or months to the largest possible group for community input. That is what seems most reasonable, anyway.--Filll 14:10, 26 September 2007 (UTC)

Filll, please see #Transclusion example for an example of one proposal, and comment if you feel it's warranted. This is only one example of possibilities being explored. Several others have put forth a couple other examples (I think), but this page changes so much from hour to hour, things rapidly get 'lost in the shuffle'. This is also one of the problems with providing examples. The example quickly gets lost from all of the newer posts and coments. Any ideas on how we can provide a consistent and easy to find place for the examples? Maybe a sub-page of this talk page with a (more or less) permanent link at the top, similar to what gets done for archives? wbfergus Talk 14:25, 26 September 2007 (UTC)
I know I volunteered to write up a sandbox, stating the case for my move/replace proposal... but 1) real life limited my wiki-time over the last few days, and 2) the discussion moved beyond my proposal... so I have held off. Since I expect that real life is not going to get slower in the next few weeks, I will have to leave stating the case for incorporating some form of change to others. It might help if everyone could make their cases more explicit and detailed. Perhaps the sand box could contain a coherent outline of both the reasons to change the wording of the section and the benefits of leaving it the same. I can understand why people who care about this policy, but have not active on the page for the last month or so, are wondering why everyone is "all fired up" to change things... and I also understand why others (who also care about this policy), are wondering why there is all this resistance to dealing with what they see as legitimate concerns. Both sides in this debate could benefit from taking a step back and explaining their positions instead of insisting on them. Blueboar 14:48, 26 September 2007 (UTC)
I still think your proposal is the closest we've come to consensus. At least, it's the only one that has gained positive comments from both the "primary/secondary" people and the "plain language" people. COGDEN 18:32, 26 September 2007 (UTC)

Everybody needs to cool off

I think that a cool-off period is needed. I am not interested in discussing motives, or diverging into polemics about policymaking, the role of the "admin cabal" in these pages, and all other miscellanea. I am taking a break from this page for a few days. ≈ jossi ≈ (talk) 14:19, 26 September 2007 (UTC)

A sensible point, and I trust that your silence for a few days will not be taken by others as assent to whatever new consensus is claimed to emerge. .. dave souza, talk 14:24, 26 September 2007 (UTC)
Obviously, others will need to step into Jossi's shoes for a bit to ensure that the basic principles of NOR and PSTS are not tampered with. •Jim62sch• 16:47, 26 September 2007 (UTC)
Nobody needs to be stepping into anyone's shoes and nobody is trying to subvert any basic principles as far as I am aware. The most telling point is that all the discussion has been on talk pages, there has been no attempt to force a change into policy. Please do not continue to insinuate that those seeking change are trying to subvert a reasonable process. Will you please at least give a semblance of assuming good faith until there are actual changes that occur to allow you to prove otherwise. I am sure we can agree that the tag edit was most inappropriate by all parties. This rhetoric is most unhelpful to a calm and friendly atmosphere which will allow us to discuss the issues sensibly. Spenny 16:58, 26 September 2007 (UTC)
Maybe one step in the direction toward consensus is to stop referring to "PSTS" as an initialism, on par with NPOV, NOR, RS, and V. The primary-secondar-tertiary source model is just one expression of one aspect of the NOR policy. Referring to it as an initialism implies that it is a policy, rather than a particular (perhaps imperfect) model of describing Wikipedia practice, and that it can't be messed with. We're going to have to "mess with" the model if we want to achieve consensus. There's no other way. The OR policy itself, and how it is effectively applied, will not change, but we have to find a way to describe the intent of the primary-secondary model without using those concepts. COGDEN 18:45, 26 September 2007 (UTC)

(outdent) I will be taking a break as well. My ability to remain civil in the face of continuing insults and bad faith is severely strained. I came to this page today ready to dive in, provide some proposals and discussion. However, I cannot do so in a continuing hostile climate without responding quite bluntly in kind, at least for the moment. My apologies. Vassyana 18:08, 26 September 2007 (UTC)

I hope you and User:Jossi will continue to monitor the discussion, because I think we'll need your agreement on the consensus replacement language for the controversial section. COGDEN 18:45, 26 September 2007 (UTC)
Heck... we need their participation to even reach a consensus that there is a controvercial seciton. But seriously... I echo what Cogden says... I hope both of you do come back once your batteries are recharged. Your opinions are valued. Blueboar 18:53, 26 September 2007 (UTC)
Seconded (if I am allowed to say that!) Spenny 19:15, 26 September 2007 (UTC)
Agree. .. dave souza, talk 19:20, 26 September 2007 (UTC)
Like an AOLer, me too... you both represent some of the most sensible reasoning and discussion that this debate has seen (along with some others). Please, take all the break you think you need, but I think we (the whole discussion, both/all 'sides') are better of with you than without you. SamBC(talk) 22:06, 26 September 2007 (UTC)
Rest assured that no decision will be made during your absences. Hopefully, by the time you return though, we might have added enough information to the Sandbox pages to continue having worthwhile discussions that actually address peoples concerns with the policy (all sides), and can begin to work on addressing those concerns. wbfergus Talk 22:13, 26 September 2007 (UTC)

What are people afraid of?

Jim, you talk about the need to "ensure that the basic principles of NOR and PSTS are not tampered with". I don't understand this comment. What basic principles do you see being tampered with? As near as I can tell, none of the various proposals that have been discussed change the basic priciples one iota. The wording that expresses those prinicples yes, but not the principals themselves. As I see it, this is not an "attack" on the principles of NOR... I see it as a discussion about the best way to express those principles. If you see the issue differently, please explain. Your comments seem to go along with Filll's comments about "agendas" and "ulterior motives"... It is obvious that you fear something will happen if we change anything ... that we will open the door to something you object to.... but what exactly is it that you fear will happen? Blueboar 17:18, 26 September 2007 (UTC)

I think that should be fairly obvious. However, let's just say we do not want anything with unanticipated consequences and other assorted blowback to happen to the project, shall we?--Filll —Preceding signed but undated comment was added at 17:28, 26 September 2007 (UTC)
It is entirely un-obvious to me. This innuendo must stop. It would do far less harm if you could describe exactly what you suspect the agenda is and explain exactly who you see pursuing this. It has been made clear that a significant number of editors are offended by your comments which they feel is directed personally at them and yet you feel entitled to carry on in this uncivil manner, presumably under the guise of a defender of the faith. I am afraid that this is exactly the behaviour that brings the administration of Wikipedia into disrepute. Would you please consider the civility guidelines and think about whether you are abiding by their spirit. Spenny 18:43, 26 September 2007 (UTC)
To repeat, Exact current wording or any proposed wording does play a pivotal rôle in the interpretation of the policy. People have known this for millennia, and hence our laws are writen very precisely (and appear to be written rather densely to many). What may seem as a minor change can have a significant impact in the interpretation and administration of a policy, hence changes must be carefully weighed and then brought to a much wider audience than is present on this talk page. •Jim62sch• 17:35, 26 September 2007 (UTC)
Neither of these points is an argument against any editing of the policy page, surely? No-one was in a rush to change anything, changes were simply being discussed. Is discussing change dangerous in itself? SamBC(talk) 17:52, 26 September 2007 (UTC)
hence changes must be carefully weighed and then brought to a much wider audience than is present on this talk page... I completely agree! That is why I ask you to explain what you fear will happen if we make a change... it is part of that carefull weighing. So what is it about the current wording that you think is vital? What is it in the various proposals that you disagree with. engage in the discussion... don't just take a "NO CHANGES!" stance and then go mute. TALK to us. What is wrong with the various proposals made so far? Blueboar 18:13, 26 September 2007 (UTC)
I was under the most strong impression that the discussion of this issue for over a month with no attempt to implement it onto the main page was a fairly good indicator of trying to carefully weigh the policy changes. I was also under the impression that the notifying the discussion (not any implemented change) widely was evidence of attempting to bring it to a wider audience. That the talk page has been obfusticated by obstructive rather than constructive comment is not necessarily the fault of the party wishing change. No doubt these points could be more concisely, but when faced with stonewalling and FUD it means that things have to be discussed in terms of rebuttals. Anyone with some exposure to legal processes will understand that to make a point is simple, to rebut a point requires detailed evidence out of proportion to the point. Is simplifying NOR really such a fearsome prospect that we might as well switch off the servers and go home? Seriously.
I'd be interested to know if you could sign up to either of my first two goals or whether you see something intrinsically harmful in the concept of having policy for the people by the people? Spenny 18:32, 26 September 2007 (UTC)
I'd be interested if someone could please produce a well documented proposal for discussion, which addresses the points raised and gives diffs of the alleged problems arising from the existing formulation. What are people afraid is being lost by the present restriction on using primary sources (as defined) for anything other than obvious fact? .. dave souza, talk 19:17, 26 September 2007 (UTC)
I asked first. :)
However, the point you have raised is entirely misplaced. What are people afraid is being lost by the present restriction on using primary sources (as defined) for anything other than obvious fact? Firstly, there has been a move in the past, which has in part been resisted by a recent edit to the main policy page, to try and enforce an outright ban on primary sources.
Secondly, the point on this is that the problem you raise is not in it being a primary source, it is the improper use of a fact to generate an analysis. If the fact was stated in a secondary source without any supporting analysis, it would not make it any more usable, would it? We are looking for a way to express the same restriction without what some (me!) see as an unnecessary mechanism of source typing. Wouldn't it be great if it didn't require Joe Public to bother with this strange source typing business yet we could still get the same result? So what if the powers that be think source typing is a useful tool? They are welcome to use that approach to divine faulty reasoning in articles. However, it has been claimed it is essential, and that claim simply does not hold up to logical reasoning.
But is a source an entire source or is a source some discrete element of a source article?
If it is a discrete element, then standard definitions of sourcing don't really help us.
If it is a whole source then we have the "it can be both at the same time" argument.
Then we have the inability to decide if certain documents or elements are primary or secondary, is it the source or is it the originality of the concept that is being tested?
Then, even when we have determined what type of source we have, we still need to decide if the sense of the edit is contained within the sense of the source. If it is primary, was the edit a statement of fact? If it was a statement of fact, was it truly represented. If it was secondary, and it was an analysis, was the analysis accurately portrayed in the edit. Those questions remain the same regardless of source typing.
If we allow secondary sources that are not of the highest repute, how do we allow this policy to let editors make reasoned evaluations of different sources to ensure that the reliability and accuracy of Wikipedia is the fairest and best representation of source material that can be achieved? If USA Today say Bush won the Florida election but the publishers of the poll said it was Reagan (I'm not very good at American politics) are we forced to go with USA Today as it is a secondary source and therefore preferred to the publishers who are bound by electoral law?
So my point is, you raise a simple point of concern. We share the same concern in principle. We want to express that concern in a straightforward way that cannot be subverted by unneeded extra steps in deciding whether something written in Wikipedia has its roots in something previously published elsewhere.
Have I said anything entirely unreasonable here? Spenny 19:59, 26 September 2007 (UTC)
That's certainly an elaborate scenario, and you seem to be in a bit of a fankle. Firstly, if there has been a move in the past to try and enforce an outright ban on primary sources, that's not my understanding and not what's said now. We should rely on third party sources, primary sources can be used with care for obvious facts. Secondly, if the fact was stated in a secondary source without any supporting analysis, then we don't go beyond the source and can't cite that source for analysis.Then, editors have to apply care and judgement in using sources. The present formulation is clear and conveys the point well to Joe Public, though some academics allegedly have problems with their specialist definitions. If USA Today say Bush won the Florida election but the publishers of the poll say that Al Gore had a majority (could this possibly happen? ;) we use a A simple formulation. ... dave souza, talk 06:19, 27 September 2007 (UTC)

I just want to note that I think Dave just summed it up very well. For me and I believe many others, what is at stake is preserving this part of the policy: "the present restriction on using primary sources (as defined) for anything other than obvious fact." As some have pointed out, this all started with an atempt o remove a line that secondary sources are prefered. Well, maybe that contested sentence could be beter-phrased, or completely rewritten. I happen to believe that PSTS is important to the article (which seems to put me in group 3) but I am also willing to revise the writing for greater precision and clarity - which either moves me to group 1, or leaves me nowhere (so to me the "three poisitions" is polarizing, I see a fourth one). Whatever further changes we make, what is most important to me is what Dave just said. Slrubenstein | Talk 19:43, 26 September 2007 (UTC)

If nobody else has done it yet, I'll go ahead and create a couple of different Sandbox pages off of this. There may a bit of trial and error getting the initial blank pages done, so please bear with me while I progress. I' will post here when I am done. wbfergus Talk 19:49, 26 September 2007 (UTC)
Okay, I have created the basic sandbox page with an example of Transclusion included in a sub-page from the Sandbox. After I post this message, I will create a link at the top of this page to the Sandbox, where other examples, essays, etc. can go. wbfergus Talk 20:21, 26 September 2007 (UTC)
Slrubenstein, what is your definition of a primary source? Is your proposal to specifically define as primary sources any category of documents that should not be cited except for "obvious facts"? If so, how granular do we need to define primary sources? Do we say that "peer-reviewed journal articles are primary sources except when they contain less than half original data and greater than half theoretical analysis"? That "interviews recorded by journalists are primary sources unless they contain at least half commentary or paraphrase"? "Movies are primary sources in the science fiction and action genres, but not documentaries, unless the documentary is at least 50% fictionalized"? COGDEN 22:18, 26 September 2007 (UTC)

Why are yoiu asking me a question when (1) I have answered it several times already on these pages and (2) expressed my willingness to modify my definition i.e. work on improving it with others, in the spirit of cooperation? What is the purpose of your question given that you know the answer? Slrubenstein | Talk 02:11, 27 September 2007 (UTC)

Wow. Well, what I read here is clearly a sign of some trouble.--Filll 22:30, 26 September 2007 (UTC)
Please could you be specific? This unspecific innuendo is very tiresome and specifically outside the civility policy. Much though I detest the quoting of behaviour guidelines myself, it should not need explanation, I will quote one specific element: Wikipedians define incivility roughly as personally targeted behaviour that causes an atmosphere of greater conflict and stress. I would ask you to consider whether your mode of conversation here falls within that definition. For the record, as a significant contributor to this section, I feel that comment is directly pointed at me. Feel free to explain to me what in my summary is an unreasonable opinion to hold and could be construed as malicious or disruptive. Spenny 22:52, 26 September 2007 (UTC)

Holy cow. I do not think I have said anything uncivil. I did not attack anyone. I have been repeatedly jumped on for my very minor contributions to this page. Is it appropriate to drive away all others from this page? All I want is, for example,

  • to see examples of what the proposed changes are
  • to see the advantages and disadvantages of each proposed change
  • to see copious examples of these proposed changes and how they would be expected to work in practice. I would like to see what would be allowed under the old policy and would be allowed under the new policy.

These should be on a sandbox page, or a few sandbox pages, for inspection by everyone.

After people spending an almost infinite amount of energy here railing away at each other, as far as I know, these do not even exist. Am I incorrect? When I express mild misgivings about how this is being done and what people want to achieve by this proposed change, it is met with very negative responses. This in itself is a very bad sign, as far as I am concerned. I am not sure what to think. Does this violate WP:NPA to say?--Filll 23:27, 26 September 2007 (UTC)

Thanks, I've no problem at all with that reply. Unfortunately, the discussions have spun out of control and there is a need to regroup. I think the issue in the recent exchanges is that there are a group of well meaning people who have seen a problem and want to do their thing to fix it. What was believed to be a straight-forward process of discuss, edit, advertise and implement, which I have seen work quite satisfactorily elsewhere (or even BRD) is now inadequate and there hasn't been enough constructive guidance on how to resolve that. Unfortunately, these exercises have been conducted within the talk page, and nobody really saw fit to suggest over the past month that there was a better mechanism to use. It is now being suggested, and is now being adopted, but at the moment it is not fit for purpose.
My observation is that the various parties have stated several times what their aims are. The response has been fairly critical for even attempting it even though I would say that we were still kicking around at the proof of concept stage. If the concerns are simply that you need confidence in a proper proposal being worked through then that is good, but it is not how the issue has been previously expressed. I think perhaps you got caught in the cross-fire, but there has been a lot of "you are doing this all wrong", a fair smattering of "We don't care what it is, don't change anything" and not enough "this is the best way to approach it." As far as the comment above, I am not making a proposal, I am simply trying to establish if anyone acknowledges that there is anything other than malicious intent.
Bear in mind that in the past very significant changes have been made to these pages without such a process so it is understandable why there is a bit of confusion as to why the bar has been raised. Regardless, various people are working towards providing that. It's bound to take some time.
Perhaps next time some innocents come along with a well-intentioned suggestion for revision, a sensible response from the community would be to help guide through an acceptable process rather than to dismiss them. If the amount of effort to make a change is to be so high when other editors have been given free reign in the past, then the least the community should do is ensure that contributors are guided in this unwritten process. Spenny 00:15, 27 September 2007 (UTC)

Where can we put stuff so it won't be lost?

I understand that sandboxes are appropriate for proposals, but would an essay at Wikipedia: PSTS concerns and proposals be a better solution? We could keep the basic summary, list of proposed solutions, etc. on the essay page, so it doesn't disappear as the talk page grows. Jacob Haller 20:13, 26 September 2007 (UTC)

Published Original Research

The guidelines seem clear than UNpublished original research is NOT allowed, but what about PUBLISHED original research? If an article I have written based on my original research is PUBLISHED in a peer-reviewed periodical, can it be included in Wikipedia? —Preceding unsigned comment added by Breadmanpaul (talkcontribs) 05:24, 27 September 2007 (UTC)

Please read the policy (look up, and click on the "project page" tab) - which answers your question. Slrubenstein | Talk 05:47, 27 September 2007 (UTC)
In particular, see Wikipedia:No original research#Citing oneself. ... dave souza, talk 05:55, 27 September 2007 (UTC)

Source-based research

As a more diverse group of editors were drawn to Wikipedia it became clear that other topics besides physics, such as politics and religion, were attracting original research, and the community sought a more systematic way to define original research and to guide editors in avoiding it. These efforts focused on distinguishing between

  • different kinds of sources and materials and
  • different ways of using sources and materials

Reliable sources

Any statement that is challenged or likely to be challenged must be accompanied by a reliable source. A statement that counts as "original research" within the meaning of this policy is a statement for which no reliable source can be found and which is therefore believed to be the original thought of the Wikipedian who added it. The only way to show that your work is not original research is to produce a reliable published source that advances the same claims or makes the same argument as you.

In general, the most reliable sources are books, journals, magazines, and mainstream newspapers; published by university presses or known publishing houses. As a rule of thumb, the more people engaged in checking facts, analyzing legal issues, and scrutinizing the writing, the more reliable the publication. Material that is self-published, whether on paper or online, is generally not regarded as reliable, but see Wikipedia:Verifiability for exceptions.

Research that consists of collecting and organizing material from reliable, verifiable 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. In order to clarify the distinction between acceptable source-based research, and pohibited original research, Wikipedia distinguishes between three kinds of source materials.

Primary, secondary, and tertiary materials

Most succinctly,

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

More specifically:

  • Primary materials include archeological artifacts; photographs; historical documents; personal diaries, census results, video or transcripts of surveillance, public hearings, trials, or interviews; raw tabulations 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).

Our policy: A Wikipedia article or section of an article can employ primary materials only if the material 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 analytic, synthetic, interpretive, explanatory, or evaluative claims. Contributors drawing on primary materials should be careful to comply with both conditions.

  • Secondary materials are analytic, synthetic, interpretive, explanatory, or evaluative claims that are based upon primary materials. It is understood that some sources may contain both primary and secondary materials.

Our policy: Wikipedia articles can include analytic, synthetic, interpretive, explanatory, or evaluative claims (1) only if such claims come from a reliable, verifiable source and (2) the point of view is clearly identified and accurately represented. The conditions that apply to the use of primary materials also applies to the use of primary source material included in secondary sources.

Our policy: Wikipedia strives to be a superb source of tertiary material. Since Wikipedia does not want to be derivative or duplicate other encyclopedias, tertiary materials are often of limited value for Wikipedia research. Annual Reviews and Encyclopedia Brittanica articles often provide extensive bibliographies that are valuable tools for identifying important secondary materials, and therefore of great use to Wikipedia editors. Nevertheless, these and other tertiary materials do not necessarily have the same content policies as Wikipedia and for this reason should not be viewed as authoritative. However, some encyclopedias and other tertiary material, such as Annual Reviews, have signed articles, and often articles that explicitly promote the author's own views. In this sense, tertiary material can also be viewed and treated as a secondary material.

Draft as of 9/20/07 (minor update)

Citing the Right 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 the right sources to back those different types of statements. Statements of fact should be cited to reliable sources that clearly demonstrate that fact. Statements of interpretation, analysis or conclusion should be cited to reliable sources that contain the same interpretation, analysis or conclusion.

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.

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 of where he was born or its effect on his career would require separate citation, since it is not an obvious or non-controversial consequence of these easily verified facts.

If we remove PSTS from this article, where will (should) they go?

Almost all of the discussion the last month or so has been around the "Sources" issue. Some of the people just joining this discussion in the last few days seem to think that this discussion is being centered around diluting NOR by merely replacing the current PSTS section section with something new, and not by actually moving it somewhere else. I think we can make progress faster on replacing PSTS if we first can agree on where and how WP:PSTS would get moved to. This would have to take place before we could ever hope to replace PSTS, as it would have to exist at least in it's entirety someplace first, or else we would be diluting NOR policy. Can we start a discussion about where and how to "re-position" WP:PSTS in either another article or in it's own "guideline" (I don't think we actually need a policy to define the types of sources do we)? Once this first step is accomplished, others can see that policies aren't being dilluted or otherwise weakened, and then many of the arguments against these proposed changes may very well disappear. wbfergus Talk 16:46, 22 September 2007 (UTC)

I think this is a reasonable question. Firstly, this is not an article, but a project page on policy. Secondly, what has been chosen here to be termed PSTS is not going anywhere. Its predecessor (Primary and secondary sources, or if you prefer, PSS) has long and extremely strong, stable widespread, community consensus. The expansion of PSS into PSTS last October has become a valuable part of the WP:NOR policy and has achieved and enjoyed strong, stable, widespread community consensus, except for some complaining here among several who do not like it, and who would prefer to totally rewrite the longstanding approach. No reasonable justification has been given thus far for its removal. Users wikiwide have come to rely on the primary, secondary, tertiary distinction in explaning and implementing WP:NOR. That section is not going anywhere on the justifications given by the advocates here thus far. ... Kenosis 17:05, 22 September 2007 (UTC)
After some limited looking around, trying to see where an existing page might be most appropriate, to me it seems like WP:PSTS could easily fit into the WP:RS guideline. It seems the most logical place, but any policy referring to it would probably need some additional words like "The following guideline section is include in this policy by reference" or something like that. A simple sentence or two like this could be added to each policy so that there is an inherent enforcement of the guideline as part of the policy text, and they would all use the exact same "text" (by reference to the guideline). This would eliminate much confusion, and when disagreements arise about types of sources, etc., those discussions would occur on that page, not scattered across the various policy pages. wbfergus Talk 17:10, 22 September 2007 (UTC)
This is presumptive and unwarranted, at least based up the arguments and weight of opinion expressed on this page thus far. Moreover, the proposal to remove WP:PSTS is a proposal to remove a long-developed integral part of a policy page, without anywhere near adequate consensus to do so. ... Kenosis 17:19, 22 September 2007 (UTC)
I just took a quick look at the talk page of each of the 4 main content policies. All four of them currently have discussions about sources. This approach of each policy having source discussions and defintions is more effective than having one page that discusses sources? If we "promoted" WP:RS to a policy in and of itself, wouldn't that centralize all arguements about sources? Future discussions on sources could then focus on the applicability of the source as it pertains to all four of the content policies, without each content policy having it's own internal debate on how that particular source issue pertains to that policy. wbfergus Talk 17:38, 22 September 2007 (UTC)
Note:This comment immediately below was intended to respond to the question farther below. I leave it here so as not to created additional confusion, since there have been responses place already.,,, Kenosis 20:24, 22 September 2007 (UTC)
In a word, no. The now infamous WP:A fiasco is a perfect example of why a focus group should not attempt to substantially change foundational policies in Wikipedia. WP:A, or WP:ATT or WP:Attribution was an attempt to consolidate what was perceived by a focus group roughly ten times larger than the focus group that's made this proposal regarding WP:PSTS. It was raised to the status of policy in February 2007, resulted in widespread discontinuity and confusion across the wiki. By the beginning of April 2007, it had been reduced once again to a proposed policy, and by the end of June was reduced further to a summary essay. The foundation of the wiki relies heavily on the balance, interaction and tensions among and between WP:NPOV, WP:VER and WP:NOR, three basic policies that are to be mediated by consensus at the local article level. The solution to the "problems" mentioned here and there in the preceding talk threads (of which there are relatively very few, anecdotal examples) is to use the existing policies at the local article level, not mess with these established policies in any substantial way. ... Kenosis 17:34, 22 September 2007 (UTC)
Kenosis, WP:ATT failed because it failed to achieve a clear consensus, which is required for policy pages. If there is a lack of clear consensus for a portion of policy, we must deprecate it, just as WP:ATT was deprecated. Your arguments do more to support the removal of WP:PSTS from this policy, than to support its continued retention. A change to something else will have to meet a standard of clear consensus. However, deprecation (as you so well point out) is a matter of simply lacking that clear consensus, not showing a clear consensus against as you'd seem to demand. Vassyana 17:43, 22 September 2007 (UTC)
That is flat-out incorrect. It achieved a consensus among the participants in the proposed policy page. Then it was rejected when exposed to a much wider consensus upon being put into play. ... Kenosis 17:59, 22 September 2007 (UTC)
That is not mutually exclusive with my point. Vassyana 18:03, 22 September 2007 (UTC)
Agreed. You just stated that it did not achieve concensus, only local approval. wbfergus Talk 18:04, 22 September 2007 (UTC)
I would not support removing WP:PSTS in the strongest possible terms. FeloniousMonk 17:15, 22 September 2007 (UTC)
May I ask why? If we can find a suitable home for it, where it is central to each of the four content policies, can you explain how placing it there would weaken or otherwise destabilize the four policies from their current form? Right now, the 4 policies have four different versions covering the exact same subject. Wouldn't making a cnetralized definition that they all use actually strengthen the existing four policies and ensure greater consistency across the application of policies? If not, can you please explain how it wouldn't and how having four different explanations of the same subject is less confusing and allows for more consistent application of policy? Thank you. wbfergus Talk 17:22, 22 September 2007 (UTC)
If I may add to this, even if people would argue that each policy needs to define things in subtly specific ways for the specific purpose of that policy, this is likely to lead to terminology clash and confusion when it comes to applying those policies together on actual articles. SamBC(talk) 18:09, 22 September 2007 (UTC)

Those interested in this discussion may wish to review the essays Wikipedia:Classification of sources and Wikipedia:Using sources. Vassyana 18:02, 22 September 2007 (UTC)

In the case of Wikipedia:Classification of sources, do please review the talk page and/or history as well, as development pettered out, so the page isn't in a particularly consistent or stable state representing any consensus of anyone. It's the idea of the page and the structure, rather than the content, which is relevant. Also, note that it was originally tagged as a proposal and was summarily re-tagged for not proposing anything "actionable". SamBC(talk) 18:11, 22 September 2007 (UTC)

Regarding the issue of where to move these - the only logical option is "to another policy page". Since they are an integral part of policy, you can't remove them to a guideline or essay (as seems to be what's being suggested here) without downgrading the status of this information (which is, of course, not an option, since we don't change policy without a complete, project-wide discussion...and even that may not be good enough). Of course, I'm still trying to figure out the "why" part of the question of moving this material. Guettarda 19:10, 22 September 2007 (UTC)

I completely disagree with this position. WP:PSTS is not an integral part of this policy, and it's not even clear why it's here, other than to attach it to a policy page rather than a guideline like WP:RS where it would otherwise be expected to be. Based on the wide disagreement on the basic definitions of PSTS (e.g. scientific vs. historical context) I would say it needs to be moved to an essay page until the definitions are revised to make them widely applicable and understandable. Then, when and if that happens, it can be integrated with the RS guideline which also needs work, and then the combined document could be considered for elevation to policy. Keeping this flawed section in this policy is only preventing needed improvement in the basic expression of the concept of PSTS, because of the inertia of policy pages. And it's preventing basic improvement in the NOR policy it is attached to, to the point of fueling an edit war that has kept this page protected for far too long. If PSTS does not have broad consensus support now, regardless of what may or may not have been deemed consensus in the past, it needs to go. All we need to do is simply define NOR without using the PSTS terms that are so confusing, by simply using descriptive adjectives that are plainly understood. Dhaluza 21:01, 22 September 2007 (UTC)
It is simple. We need to distinguish between PSTS because distinguishng between verifible and non-verifiable has not been sufficient to explain the diferrence between OR and non-OR.Slrubenstein | Talk 22:03, 22 September 2007 (UTC)

It was once the case that WP:NOR was more simply stated-- about three or three-and-a-half years ago. The current form of WP:NOR is a result of that three-and-a-half additional years of practical experience, during which time WP has grown from something like 100,000 articles to over 2,000,000 articles. Users around the wiki, myself included, constantly run into the need to explain the WP:PSTS distinction w.r.t. WP:NOR. Reducing this policy to a bare-bones version again based upon vague complaints simply will not wash, at least not without a clear, widely consensused improvement that takes into consideration the same range of issues that led to its current form. There may well be more effective ways of doing it. But we certainly have not seen them articulated here in any way that factors in the breadth of the issues that people encounter in the numerous types of material and topics areas covered in those two-million-plus articles. ... Kenosis 22:24, 22 September 2007 (UTC)

Naturally NOR has changed over the years, but that is a reason why it should continue to change, not a rational for no further change. The fact that you have to explain PSTS should be a sign that change might be warranted. I don't think that there is serous discussion about stripping the policy down, but there is serious discussion of the cost/benefit of PSTS, and if it does not have wide consensus support, it must go regardless. So, can you please explain for my benefit why we need to use wiki-defined terms like PSTS, rather than plain English, to describe NOR? Dhaluza 23:34, 22 September 2007 (UTC)

Arguments agianst moving PSTS anywhere else

Okay, so far the arguments against moving WP:PSTS seem to be the following:

  • WP:ATT failed, because of a local consensus that did not withstand broader consensus. This is the destined situation for moving or removing WP:PSTS.
  • The current Wikipedia four main content policies each having their own section about sources, defined and presented in differing contexts. This is considered important as each policy has a different focus and context.
  • Each of the four main content policies needs to define "sources" differently because each definition is an inherent part of each policy. For instance, in order to best ensure there is no original research in an article, it is imperative to understand primary or secondary sources.
  • Deprecation or replacement of this section of policy would have negative consequences, including a weakening or destabilization of policy.

Did I miss anything? wbfergus Talk 18:03, 22 September 2007 (UTC) Revised 18:40, 22 September 2007 (UTC) by Vassyana.

I agree with your essential points, but this a bit snarky. Please consider rephrasing. Vassyana 18:06, 22 September 2007 (UTC)
I agree with Vassyana. I'm not actually sure how to get the same essential point across in a better way, mind you, but I agree that some way ought to be found. The main point is to ask if those opposing feel that this reasonably summarises the arguments, what's included that shouldn't be, and what's missing. SamBC(talk) 18:15, 22 September 2007 (UTC)
I agree that I was probably "over-the-top" with my sarcasm, and I apoligize to any who may be offended by it. But, I am also at a loss as to how it could be re-worded unsarcastically. I've gone over the arguments against any change at all, and this seems (to me) to what the main points against the proposed change are. When asked for specific examples of how this is bad, these same points come up again and again. If anybody wants to reword the above any any way, please do so. All I ask, for historical purposes, is that you use strikeout instead of just replacement. I am trying to determine why some people are against the proposed change, or how this won't eventually help improve Wikipedia with consistent definitions and consistent enforcement of policy based upon those consistent definitions. Thank you. wbfergus Talk 18:15, 22 September 2007 (UTC)
I've tried revising the points to be more neutral and civil. Since I am rewriting another's comments, please revert me if it is felt inappropriate. Vassyana 18:41, 22 September 2007 (UTC)
I feel some of the points now appear weaker than how they were first presented by the original poster of the point, but I can live with the more nuetral wording. Thank you for your effort. wbfergus Talk 19:00, 22 September 2007 (UTC)
  • The representation of "arguments against" in this section is wirtten and placed by an advocate of removing this longstanding policy, and should not be taken as a summary of arguments in objection to the proposals presently being put forward on this page. No further comment at this time. ... Kenosis 18:11, 22 September 2007 (UTC)
No thank you, at least at this time. What I am beginning to see here, and above, are strategies that are arguably crossing the line into WP:POINT. I've said my piece for now, thank you. ... Kenosis 18:19, 22 September 2007 (UTC)
It seems instead that there are no valid reasons opposing the proposed change. Slrubenstein (who I have grown a great respect for during our discussions here the last month), seems to be the only person who is opposed to the proposed change who has bothered to even attempt to come up with valid points while also offering some very constructive criticism. For almost all other opposition to the proposal, when asked for specifics, we get variations of the above points. When asked how the proposal could be made better to alleviate unspoken concerns, no suggestions are forthcoming. Please, please give us examples of how this will destabilze any existing Wikipedia policy, lead to greater confusion, or make enforcement of any of these four main content policies more contentious? wbfergus Talk 18:43, 22 September 2007 (UTC)

We should not move PSTS. While Wikipedia articles on different branches of academia, journalism, and jurisprudence can have discussions of how experts in those fields conceptualize and categorize sources, for WP policy the distinction is meaningful only in the context of this policy, because the concept of "verifiable sources" is not enough to explain what original research is and we need to make other distinctions about sources in order to clarify the different ways they may and may not be used in Wikipedia. As such, any discussion of them belongs here. Let me add that the more I reflect on this, the more the whole debate seems like a wild goose chase. It typically begins with someone faulting the policy for prohibiting the use of primary sources. The problem with this argument is the policy has never flat out prohibited the use of primary sources. As soon as someone points this out, critic just start arguing that the whole mention of sources is problematic. It is just a bait and switch tactic. That said, I have no problem with trying to improve the text, and indeed I have tried to do so myself. Slrubenstein | Talk 18:19, 22 September 2007 (UTC)

I think the suggestion of fraudulent arguments is offensive and has the smell of someone not accepting that the points have been argued in good faith.
The fundamental point is a simple one: however you define a source, it can be used correctly or incorrectly. The action of OR is not dependent on the type of source, but the editing.
To go further, the text suggests that sources as a whole are primary or secondary, but is seems to be consensus that sources can be both. Then people introduce the concept of perspective, where a source changes its nature dependent on the point being made. Then we have the debate of whether the source is a cohesive whole, or a collection of concepts and each concept needs to be tested.
There is no bait and switch: there is a simple observation that the solution appears more complicated than the problem. I do problem solving on a professional basis, and I can guarantee you that the biggest cause of "impossible" problems to resolve is the insistence of educated people refusing to accept that a problem that causes difficulty can be the simplest one that has already been dismissed as it is too obvious. Source typing seems a nice pure solution, but for the purposes of NOR, especially outside the domain of good academic sources where there are still sound encyclopaedic articles to be written, it simply is not.
The edit warring over the previous month was not about removing the categorisation, but simply a genuine deadlock over being able to reconcile the categorisation with appropriate uses. The general consensus was that there had to be exceptions, and the confusion was that nobody seemed to be working to quite the same concept of primary and secondary sources so could not agree what were the real exceptions.
The classification of sources is of no interest to me: I simply fail to see how it resolves the question of "Says who?" and it fails the general public test of being able to explain simply and concisely what No Original Research means. If Wikipedia is the domain of a few academic wizards and everyone else are inconsequential ticks to be looked down on and deemed a nuisance, fine, let policy owners play the academic game and come up with complex rules that Joe Public will never really understand. However, that is not the domain Wikipedia occupies at the moment, and therefore policy should be set out in the simplest and most accessible way that it can be. Spenny 21:24, 22 September 2007 (UTC)
Once again, I will ask that those opposed to this proposed change please do the following:
Open each of these in a new window and compare:
Now, after doing this, please describe how this is more efficient or leads to less confusion than if all of these various definitions and examples were instead placed on one 'page' somewhere, with each of those policies mentioned above linking to it with the following statement (or one similar in concept), "The following guideline (or policy, whatever it turns out to be) is also included in this policy by inference". The referred page could even have subsections in it for specific examples of what or how "Sources" directly related to each policy. Having everything in one place for a "subject" like sources easily allows everyone to quickly view all of the information in one viewing. The 4 different policies are currently attempting (very poorly in my opinion) to piecemeal this subject, instead of trying to improve peoples comprehension of the subject. It doesn't appear at all that any of those opposed to the proposed change have done this, as none of those have expressed how this current arrangement is better than what is being proposed. wbfergus Talk 18:34, 22 September 2007 (UTC)
What is being requested here is a summary of the entire history of the three core foundational editorial policies upon which Wikipedia is built. The onus is not on users to defend the entire history of existing WP policy, but rather, the onus, or burden of proof, is on those who assert that significant changes should occur. As to the specific proposal to alter, move, or remove WP:PSTS, I repeat here again what I gave above:

...... The section on Primary, secondary and tertiary sources is a vital section that puts WP:NOR into perspective for users around the wiki. The proposal to replace the section or move it elsewhere is a major change in this policy page that I find unacceptable, at least without something that represents a clear improvement on it. Since well over three years ago, a distinction had been made between primary and secondary sources and what uses constituted original research or original synthesis. A significant number of WP users remained confused about the distinction, e.g., "well, if WP is not supposed to synthesise material directly from primary sources (except for purely factual matters), and if we're supposed to use secondary sources, then how is WP a secondary source?" The extra note about tertiary sources was inserted into WP:NOR nearly a year ago, and suddenly it made better sense to most other participants and much of the confusion diminished substantially. The description of encyclopedic matter as primarily tertiary helped to resolve this confusion right on the policy project page. Since then, it's tended to serve us well in resolving confusion about what is original synthesis, or original research, as distinguished from original wording and editorial decisionmaking.

...... Admittedly there remain residual bits of confusion. One user mentioned Encarta and Britannica, which are also "tertiary sources" according to the delineation presented in WP:NOR. I pointed out somewhere in these rapid-fire multiple threads today there is a wealth of more specialized tertiary sources such as specialized encyclopedias, to which we can also add many textbooks as tertiary sources. Indeed one could readily also argue that there are numerous sources that are quaternary, quinary, senary, septenary, octonary, etc. etc. Categorizing sources as primary, secondary and tertiary has, nonetheless, served quite well in clarifying the basic types of sources from which material in WP is drawn. If there is a proposal to better clarify the distinction or explain that the delineations between them are not always sharply drawn, or whatever, that would, as I said in the section below, be quite welcome and potentially helpful. But the proposal to replace it with the language at the top of this talk section is, IMO, completely unacceptable on the justifications given thus far. I believe I used the words "summarily reject" to covey my summary opinion about the proposal. ... Kenosis 18:49, 22 September 2007 (UTC)

I fail to see how two paragraphs above answer my question, how are the four (current) definitions in the four different core policies less confusing to the majority of Wikipedia editors or how the four different definitions lend themselves to a more consistent enforcement of policies across all of Wikipedia than just one definition that each policy used would. Answering the above question with two completely unrelated paragraphs is of no assistance, when assistance is requested. wbfergus Talk 18:56, 22 September 2007 (UTC)
The reason that it is unreasonable to expect an answer to the question just posed above was given in the first of three paragraphs I gave just above. And the failure to "see" how they relate to the request, essentially, to explain the entire justification for core WP policy is not sufficient grounds to assert the need for a significant change to longstanding foundational policy, nor to the way foundational editorial policy has been long allocated to the three most basic core editorial policies, WP:NPOV, WP:VER and WP:NOR. ... Kenosis 19:11, 22 September 2007 (UTC)
Additionally you state that "The section on Primary, secondary and tertiary sources is a vital section that puts WP:NOR into perspective for users around the wiki". Since this is also covered in the other 3 main content policies, and PSTS also links to three different pages with far more detailed explanations and examples, how you state such a claim? It can't be proven that what you say is in fact true. It also can't be proven that maybe many more Wikipedia editors received their "guidance" from one of the other pages, or maybe even from another editor. Your statement itself is based entirely upon "original research". wbfergus Talk 19:05, 22 September 2007 (UTC)
Unfortunately, it is increasingly becoming clear to me that the primary failure here may be a failure among the several outspoken advocates of policy change to understand the basic policies. But a complete, thorough rehashing and justification for the long, increasingly rich history of the core editorial polices, from their beginnings to the present, is out of the question. It's just too much to effectively do here, though it might well be a very reasonable project to embark upon for the future. ... Kenosis 19:17, 22 September 2007 (UTC)
...so, if presumably intelligent people are failing to understand the policies, is it the people or the explanation that is the problem? Spenny 21:33, 22 September 2007 (UTC)
Understanding the three core policies and how they interact and are interpreted via the consensus process takes time and work. It's never been instant, and seldom has been a quick learning process. The material presently included in WP:NOR, as well as the other two core editorial policies, is mostly a result of accumulated experience with the application of the policy. Removing WP:PSTS from WP:NOR makes the learning process more contorted and confusing in practice, not less so. ... Kenosis 22:35, 22 September 2007 (UTC)
So essentially you subscribe to belief the Wikipedia policy is far to complex for the common man to be able to intuit. Sounds more like encyclotology than encyclopaedia Spenny 22:54, 22 September 2007 (UTC)
That, sir, is a rather hyperbolic snipe to say the least. You've made your opinion clear. This page, however, is not going back to square one, not without lengthy and broad community discussion far beyond the miscellaneous and inconsistent complaints that have been intermixed here of late, and not without a proposal that can be widely agreed to be a clear improvement. The current basic structure of WP:NOR, including WP:PSTS, is based upon experience, and is presently part of the wiki method of avoiding original research or original sythesis. WP:PSTS guides the user, as an integral and long accepted part of this policy page, to better understand what's meant by WP:NOR in the many contexts in which it is manifested across the wiki. ... Kenosis 23:09, 22 September 2007 (UTC)
I was rather pleased with it myself as you might guess, but yes, a little bit too snipey, so apologies for that. However, the point was a serious one, do you not see the contradiction in what you say and the purpose of expressing policy? If you pick up [this link, I think you'll see that I am trying to avoid square one. However, I do think the long term goal should be to get back to a much simpler expression of policy over some time. Spenny 00:59, 23 September 2007 (UTC)
Spenny, essentially you seem to be ignoring the genuine concerns of experienced editors who find it very hard to see that this proposal with incompletely documented justification is going to improve things by removing a well used and generally understood part of the policy. Comparing that to Scientology sounds like an insult, and you're surely familiar with WP:EQ and WP:CIVIL. if, as is now being said, the intention is to refocus this section and move it elsewhere, there should be a proper proposal as to how that's to be done. Since it interlocks with the use of sources in a way that avoids OR then there will still be a need for a summary style outline of it in this policy. .. dave souza, talk 23:20, 22 September 2007 (UTC)
The allusion is simply that there are those who appear to claim a deeper knowledge than the ordinary mortal. With regards to civil, as I have alluded to elsewhere, it seems that if the motives of those who know is impugned, then that is incivility, when the motives of those who do not is impugned, then that is acceptable. There have been several posts on this page and elsewhere which suggest that people suggesting change are not doing this with sound motives. The problem I have with the recent discussion is the strong sense of denial, the this is how it is, anyone who does not think as we do has clearly not understood, you have not enough years of experience to commune with consensus. To be frank, that is all pretty condescending and uncivil, not assuming good faith - and I didn't need any of those nasty square brackets to say it.
I am not ignoring genuine concerns but I don't think I've yet heard a reasoned argument that refutes that there is a case to answer on this page, that case being the policy is not expressed in a simple and comprehensible form. My point has always been very simple: you do not need to descend into disputed terminology to be able to explain the principles, and the principles do not depend on typology. If you care to check back I had a stab at addressing that without major surgery to the wording. You would also see that I do actually have a concern that too much is being lost and I believe there is a middle ground (but that middle ground does not need the words primary and secondary within it). However, there is little point expending the effort to reiterate this, if there are those who are just saying that'll never fly when I know they can effectively veto change without ever engaging in discussion. Spenny 00:39, 23 September 2007 (UTC)

Ok, so I'm trying to figure out what's going on here. Can someone point me to the arguments in favour of removing this section from the policy page? Guettarda 19:06, 22 September 2007 (UTC)

For a (very) brief synopsis, please see #Attention_newcomers_to_this_article. I "think" I covered all the main points there, but I may have missed something. Feel free to ask any additional questions. wbfergus Talk 19:10, 22 September 2007 (UTC)
This proposal has come together recently on the presumption that those taking part art familiar with arguments extending back over several archives, and are convinced that "something must be done". The evidence for this has to be set out in a coherent form so that it's accessible to people who haven't been involved in this long process, and it was proposed earlier that an essay is to set out the case. The idea of removing PSTS categorisation of sources seems to be aimed at some undefined difficulties of interpretation. The effect would be to remove from this policy restraints on research from primary sources. At present this policy allows their use for undeniable facts. Other policies require use of third party sources. In my opinion there's an inconsistency there that should be clarified. While the wording can be reviewed to avoid any confusion, removing restraint on using primary sources is not acceptable. It may be possible to move detailed clarification of clarification to a resource shared between policies rather than having it here, but the essence of the policy has to remain in WP:NPOV. ... dave souza, talk 19:15, 22 September 2007 (UTC)
I don't think anyone is seriously arguing for removing the existing restraints on factual sources. We just don't need the excess baggage associated with defining them as primary sources, which means different things in different contexts. Dhaluza 21:09, 22 September 2007 (UTC)
Okay, Dave, I'm going to say this again. Please respond this time. A number of times above I have given the explanation and examples, so I'll stick to a bland statement this time. "Primary source" and "third-party source" are not mutually exclusive. Primary does not mean or imply "first-party", although one may say that it means "first-hand", but that is not the same thing. SamBC(talk) 23:25, 22 September 2007 (UTC)
Right, Wikipedia:Classification of sources is obviously giving an indication of the distinction you're drawing. The present wording in NPOV NOR defines primary sources as sources of fact and as being close to the subject: that article's putting the latter definition into the "first party" area. The first essentials that come to mine, and it's a bit past midnight so not the best time for thinking, are that facts that need analysis etc. have to be in the context of an outside source for anything more than plain description, and that analysis or opinion close to the subject similarly needs an outside view. In both cases this has to do with avoiding original research in presenting the information. This discussion needs focussed, but not tonight. Any progress yet on #should we make a proposed change essay page? ... dave souza, talk 00:01, 23 September 2007 (UTC) corrected' dave souza, talk 16:45, 23 September 2007 (UTC)
I'm going to have to guess that you made a (late-night-induced) mistake saying NPOV there, because I can't find anything about primary sources in NPOV; assuming you meant NOR, then I think part of the problem is the ambiguity of the term "close to the subject". Based on the general (including off-wiki) use of the "primary source" label, I would read that as meaning "able to make first-hand observations". After all, a witness to a traffic accident provides an account and that's primary (and I sincerely doubt that anyone would disagree with that). However, the witness is a third party unless they were actually involved in the accident (or certain other situations, like having a connection to one of the people who was). They are "close" to the subject (the accident) in a very literal sense in that they were there (or nearby) at the time. Hence, first-hand but not first-party. I'm also retiring for the night around now, when it's gone 1am I'm generally not at my best. SamBC(talk) 00:20, 23 September 2007 (UTC)
Thanks for picking up on my typo, it was after 1 BST when I posted that. :-/ Ιt's a very good point that "third party" has just as much of a problem in terms of ambiguity, and the relation of terms categorising sources in all policies needs to be straightened out. The #Motivations section below seems to be looking for a way forward... dave souza, talk 16:45, 23 September 2007 (UTC)
Reply to Slrubenstein: you said "It typically begins with someone faulting the policy for prohibiting the use of primary sources. The problem with this argument is the policy has never flat out prohibited the use of primary sources. As soon as someone points this out, critic just start arguing that the whole mention of sources is problematic. It is just a bait and switch tactic." I'm not using a bait and switch tactic. I'm concerned that some versions of the policy may be interpreted as prohibiting the use of primary sources. I agree that the policy has not flat out prohibited primary sources, but it has said in some of its versions that the use of them should be "rare" or "used sparingly" and in some versions has stated in bold type that Wikipedia articles should rely on secondary sources, which might be interpreted as implying that they should not rely on primary sources. (Currently it says articles should rely on primary and secondary sources.) My concern is that those sorts of wordings could be used, by editors interpreting the policy, to prohibit some specific uses of primary sources in some articles. I believe that if the policy says something should be "rare", for example, then it will be used to prevent the thing from happening on some pages at some times, with the effect on those pages being the same as if the policy prohibited it. This is a genuine concern, not a tactic for the purpose of promoting some other goal. My concern is particularly strong when it's not made clear in the policy that peer-reviewed articles are not to be generally lumped in with "primary sources", but my concern is also there for primary sources in general. Perhaps NOR needs to say that care must be taken when using primary sources and that one should not go beyond what the sources say, (etc.), but in my opinion it should not limit how often such sources are used. --Coppertwig 14:59, 23 September 2007 (UTC)

wbfergus, you wrote at the start of this section "Each of the four main content policies needs to define "sources" differently because each definition is an inherent part of each policy...." Did you mean three content policies WP:NPV, WP:V, WP:NOR, and the content guideline WP:RS guideline or did you mean something else? --Philip Baird Shearer 13:21, 26 September 2007 (UTC)

For the fourth, I meant the one main core content policy that is almost always overlooked, WP:BLP, probably because it doesn't always apply to every article, just the articles on living persons. wbfergus Talk 10:42, 28 September 2007 (UTC)

Goals

I was tempted this morning to archive the whole talk page and get everyone to make a fresh start. Here is my fresh start, to see if I can convey my motives for change. With anything on this, we get into a why am I bothering: all those dangerous principles that get us worked up and put far more effort into something that does not deserve the attention it is getting.

I will set down a couple of my starting points. Let's test to see if they are worthy goals. If they are, then we can work together to find a way to satisfy them. Take it as read that I do not believe the policy pages meet these goals. I would like to think that if you go back through the pages, all my comments have been based around these principles. Perhaps you have some goals. Perhaps having agreed a common purpose, we can then test all the policy pages against these principles and, slowly and carefully, work towards these goals.

1. Policy for everyone

Policy is not the domain of a select few who have communed with the Oracle. Wikipedia is a community and within some bounds, all members of the community should be welcome to discuss policy constructively. Contributions not be in keeping with the goals of a great encyclopaedia can be robustly challenged, but constructive criticism should be welcomed, as should constructive contributions.

2. Policy should be stated in straightforward terms

(a) It is a fundamental for the true acceptance of a policy that the policy is understood. It should not be written in such a way that it requires significant intellectual capacity to divine its subtle nuances. Put another way, it should say what it needs to say in the simplest way possible, but no simpler. The expression of policy is not the same thing as the policy.

(b) Wikipedia policy should avoid its own neologisms: I have felt that the primary/secondary debate has in part been fuelled because the policy is expressed in terms that can only be understood if you understand the policy.

This is my Campaign for Plain English goal. There should be no unnecessary intellectual barrier to understanding policy.

As an aside, I would set a plain English goal for the whole of Wikipedia - it might be there, but I was struck by yesterday's comment which I read as being proud that the reading age of some articles was high (that may not have been the intent).

3. Stability

I put this up, not because I agree with it, but because I strongly disagree with it when it becomes a goal in its own right. It becomes an excuse for refuting criticism. In the past, I have had this argument made to me explicitly, and it is an underlying theme. Put another way, stability is useful if we can achieve it, but it is not necessary, and if it works against (1) and (2) then it is undesirable.

Comments and concerns

Agree:

1. Comments that WP editors need higher education are elitist, and not consistent with the principle that WP is an encyclopedia that anyone can edit.
2. Policy should avoid neologisms the same as articles, should use plain English as well. Policy statements should not be candiates for {{Confusing}} {{Unclear}} or {{Technical}}.
3. I think the people rigidly defending PSTS need to consider that a similarly rigid group of editors could have prevented its inclusion in the past for the same reasons given now.

-- Dhaluza 09:54, 26 September 2007 (UTC)

  • Agree in principle. Probably the reason more people haven't been participating on this is because most users don't know this debate is going on. It seems to me (going from memory over the last month or so now), that probably 2/3 of 'contributors' to this debate have been for change, and about 1/3 against change, mostly within the last week or so. There also appears to be a small core of people in the latter group, while being against change, are open to change if it doesn't weaken policy and makes things clearer and less contentious. wbfergus Talk 10:21, 26 September 2007 (UTC)
  • These are your own goals. Looking at them,
1. Fine, but to challenge and change a policy you have to put up a coherent case well supported by evidence (not just anecdote), and welcome constructive criticism of the proposed changes. Comments that WP editors need to be able to read are hardly elitist.
That was unnecessary misrepresentation of the point. Spenny 14:56, 26 September 2007 (UTC)
2: Simple terms that many, probably most, editors can follow should not be junked because some PhD level disciplines use the terms slightly differently. If there are better terms these should be proposed and subjected to detailed critical testing to ensure that they don't change the effect of the policy. Alternatively, proposals can be considered for improving the explanation of the terms.
No argument there. Spenny 14:56, 26 September 2007 (UTC)
3: Stability is essential for contributors who make the effort to get to grips with the policy, but don't keep it on their watchlist and check it daily to see if someone's changed it and they have to begin a new struggle to get to grips with the new formulation. New terminology can be introduced if it maintains in full the effects of the policy, and if an explanation is included showing how the new terminology relates to the previous terms. .. dave souza, talk 14:19, 26 September 2007 (UTC)
I don't find this convincing, though I understand your point. There has never been any suggestion that policy is fundamentally changed, just the explanation (plus, I should add, some concern that there has been some subversion of the policy - as recent edits to resolve this were made with some consensus). On that basis, those who understand policy should not be concerned about its latest formulation, if it looks like NOR, smells like NOR, then, by golly, we've got a good idea that it is NOR. I agree that, for example, removing source typing would be a jolt, but given that I do not subscribe to the belief that "It's a primary source" is equivalent to "It's NOR" then that sort of change might be a useful one to adjust to, we can call it without bringing other terms into the main space. In other words, editors will use the policy pages to understand the key principles and it is rare that they will need to be concerned with the exact current wording which is used to convey the principles. Spenny 14:56, 26 September 2007 (UTC)
Exact current wording or any proposed wording does play a pivotal rôle in the interpretation of the policy. People have known this for millennia, and hence our laws are writen very precisely (and appear to be written rather densely to many). What may seem as a minor change can have a significant impact in the interpretation and administration of a policy, hence changes must be carefully weighed and then brought to a much wider audience than is present on this talk page. •Jim62sch• 16:54, 26 September 2007 (UTC)
Two different points: I don't see anyone disagreeing about careful wording, where do you? My point was, assuming that the changes have been carefully weighed and promulgated to the appropriate audience, the argument of monitoring policy pages for changes in the explanation is a straw man argument. We especially need to remember these are not laws but guidance and we are enjoined, where appropriate, to ignore all rules. On that last point, I have made an observation on the IAR talk page for some tightening up to avoid the presumption that it might mean what it says. Spenny 18:21, 26 September 2007 (UTC)
We have to be careful about Wikilawyering here. Policy pages are not legal codes. They are pages that try to accurately document current Wikipedia practice and convention. (See Wikipedia:Policies and guidelines.) The only true policy is the convention itself. I believe that everybody here essentially knows what the consensus OR policy is, and can apply it correctly. Where we differ is in how we describe that policy. The policy page description has to be a consensus description. The policy page is to educate, not to decree. Thus, I totally agree with points #1 and #2. I also agree with stability, but only stability of consensus policy itself. If a policy page contains a non-consensus section, no matter how long it remains there, that works against stability, because it seeks to change the long-established behavior of Wikipedians. COGDEN 18:13, 26 September 2007 (UTC)
Wikilawyering? None was offered here -- it was an analogy.
In any case, in order to describe policy properly we must be sure that the descriptions are all-encompassing and not simplified to the point of real or potential dilution of the policy. That was the point. Language is a tricky thing -- simplification often has the effect of complicating an issue. •Jim62sch• 20:20, 26 September 2007 (UTC)
I agree. But I'm not worried about policy being diluted. That's impossible in the short term. The policy exists. It has existed for years, and nothing we do or say on the policy page can change that. All we really need to worry about is describing current Wikipedia practice in a way that most everybody agrees is not incorrect. Of course, the more understandable, the better. It won't ever be perfect, but it will achieve consensus, which is something sorely lacking with the pseudo-historiographic primary/secondary model. And if we can never achieve consensus on any language in this area, the policy page will be blank. Even if it's blank, though, nothing will really change in practice, because the policy is current practice, independent of how we describe it on the policy page. COGDEN 22:04, 26 September 2007 (UTC)
Re the last sentence just above by COGDEN: No. It's interactive. Sometimes policy leads the way, and sometimes it follows the practice of experienced users that are familiar with the concerns involved in how to write and edit articles. Moreover, the three core content policies were set into motion, essentially, by decree from the founder. NPOV, VER, and NOR, were each set into motion "from the top", so to speak. From there on it developed by consensus, was honed by an increasingly large body of practical experience, and modified accordingly as it developed. At this stage, evidence of successful application of WP:NOR, including PSTS, can be found in every topic area of the wiki. So, it reflects practice and also defines it in a way that both new users and experienced users are also expected -- indeed required -- to follow. A blank page doesn't accomplish this, nor does a distilled or gutted version ... Kenosis 04:29, 27 September 2007 (UTC)
Additional note: The word "distilled" in my last sentence just above was by no means intended to advocate against occasionally distilling policy and guideline pages, because WP:CREEP is a reality in an expanding realm like WP that can get out of control, and which occasionally does get out of control. And when the "instructions" within a policy or guideline are consensused to have gotten out of control, they should be dealt with accordingly and either split off into a fork or distilled to manageable size. In my opinion, thus far this is not the case with WP:NOR w.r.t. the PSTS section presently being disputed by several users on this talk page. The potential for expansion of how PSTS is applied in various contexts throughout the increasingly large wiki is why I find the proposal by several users for a separate WP:PSTS page reasonable. In my opinion, such an additional page: (1) could, as already mentioned, be transcluded onto this page with the basic already existing content of the present section on "primary and secondary sources" left intact, or (2) could be transcluded from this page to the proposed WP:PSTS page, leaving supplementary commentary to the additional page, or (3) simply be a separate essary or guideline about how PSTS has been successfully applied in practice in various contexts in Wikipedia, or (4) could be some variation on one or more of the three I just mentioned. Removing the section on PSTS is, IMO, out of the question here, because its predecessor "primary and secondary sources" has been in play almost from the beginning of WP, and "primary, secondary and tertiary sources" (PSTS) is presently a reasonably well thought out and widely accepted expansion of the original that uses tertiary to accommodate an extra dimension of secondary source such as encyclopedias and other publications that clearly are designed to integrate divergent secondary sources. ... Kenosis 05:31, 27 September 2007 (UTC)
It sounds like you disagree with Wikipedia:Policies and Guidelines, and you'd favor changing that page to allow for consensus-based policy pages that lead, rather than follow, editors' behavior. I'd take that issue up on that page. But you'd really have to show a lack of consensus on that policy, much as we've done here. But then, showing that the Wikipedia:Policies and Guidelines page does not reflect actual Wikipedia practice and therefore must be changed would sort of defeat your argument here. Kind of a dilemma.
The primary-secondary-tertiary model has not been in play since the beginning of WP. It's been around for about a year now, mostly without significant impact, since the majority of Wikipedia sources are still primary sources. In fact, I'd say based on my own unscientific observation, that the ratio of primary to secondary sources has actually increased during that time, especially within featured articles. COGDEN 20:37, 1 October 2007 (UTC)

NOR is useful. The word "primary" is not.

Above I said:

Slrubenstein, this is the crux of the whole thing. Please give the best example (make one up) you can of a situation where the word "primary" is an asset rather than a liability in communicating the reason that you reverted someone's edit. WAS 4.250 11:28, 26 September 2007 (UTC)

The response to that and other similar statements is that people see "primary" and think "NOR". The issue here is not NOR. The issue is the word "primary". Read that again until you get it. We are all very tired of the accusations and ridicule based on what ... inability to read? Inability to AGF? No one is asking that the substance of NOR be changed. Only the words we use to explain NOR. Why? Because the word "primary" gets in the way of a clear explanation of what the violator of NOR did. Why? Because we use a Wikipedia specific definition. It is a Wiki-ism that gets in the way of a clear explanation of NOR. You violated NOR. No I did not. You used a primary source. No its not. Well it is by our definition. What is your definition? blah blah blah. Well so what. NOR does not say primary sources can never be used. Ya but you are misusing the source to make a claim the source does not make. Well for the love of God why the frickin hell didn't you say so in the first place? WAS 4.250 23:43, 26 September 2007 (UTC)

WAS, are you addressing me? Are you accusing me of ridiculing and accusing you? Because it seems to me that you are attacking me. Above both wbfergus and jossi (who, I think, disagree on other things) have urged people to be more civil to one another. If I have been uncivil to you I honestly do not recall that but just in case, I apologize right now. But I think if you reread what you just wrote after allowing some time to pass you might see why it seems to me that you are just calling me a stupid point-of-view pusher. Are you not satisfied that i have expressed myself clearly? Fine, that is a fair point to make and you can ask me for clarification - but it really seems to me like you are simply asserting that I lack good faith here. That is not fair.
I also do not understand your hysterical "Why for the love of God why the frickin hell ..." I have never - never ever ever - said that primary sources cannot be used. Indeed, i have consistently said over the past three plus years that primary sources can be used under certain conditions. I repeated it, for the zillionth time, in the section above, in response to a comment by David souza. This is not the first time I have said this; there is no "finally!!!!" - I have stated this repeatedly as long as there has been an NOR policy. My intention above was merely to thank David for expressing an important point so clearly. But you feel some compulsion to turn my expression of Wiki-love into an excuse for you to spew wiki-hate? Now will you please explain to me why you want to attack and insult me? —Preceding unsigned comment added by Slrubenstein (talkcontribs) 02:47, 27 September 2007 (UTC)
I have to second the sentiments of Slrubenstein here. People, please calm down. This is really a tempest in a teapot here.--Filll 03:18, 27 September 2007 (UTC)
Agreed it was unhelpful to express it in those terms in the current atmosphere. WAS does get misinterpreted. Please note that the main thrust of the point was to mimic a discussion with a user, not aimed at a person. I've picked up the theme below. Spenny 08:32, 27 September 2007 (UTC)
I don't think that was intended as a personal attack. I think that was intended as a slightly humorous narrative of an edit dispute. I can recognize the arguments and the mutual frustration. The last part is an expression of the frustration many editors have felt in debates over whether statements are appropriately sourced, appropriately balanced, etc. Jacob Haller 03:01, 27 September 2007 (UTC)
I am glad I misunderstood, then! Slrubenstein | Talk 08:40, 27 September 2007 (UTC)
Now, to answer your question. This map was included in the White people article:
 
Hair color map according to Frost [1]. The yellow represents 80%+ light hair, orange is 50-79% light hair, tan is 20-49% light hair, dark brown is 1-19% light hair.
 
Eye color map according to Frost [2]. The purple represents 80%+ light eyes, green is 50-79% light eye, pink is 20-49% light eye, dark brown is 1-19% light eyes. Grey area omitted by Frost.
I would call it a primary source, or, if some people quibble with the word choice, primary material. By my definition it is primary because it presents facts (the distribution of eye-color in Europe) and - by my understanding of NOR, if someone presented this solely to illustrate the distribution of eye-color in Europe I would consider that perfectly acceptable under our NOR policy as currently written. I also consider this to be a primary source in that it provides data which can be and has been the object of analysis. Some of the explanations for this distribution has to do with geography and selective pressures acting on eye-color at different latitudes. Some of the explanations have to do with the history of migrations and intermarriage among neighboring groups. These explanations constitute secondary sources or material, and they have been forwarded by a variety of scientists, and if an editor added "Scientist X has used this distribution to illustrate her point about latitude and selection" or "According to scientist Y, one explanation of this distribution is a postulated migration of people ..." I would also consider these perfectly acceptable under (even desirable and to be encouraged by) our NOR policy as currently written. However, an editor inserted this map in an article to illustrate that people of Northern Europe belong to the White race, because one of the defining features of the White race is a tendency towards blue eyes. Now, I consider this a violation of NOR because it is taking the primary source material out of context and using it to forward the editor's own point of view. This is especially troubling because many scientists uses this data to explain why they have abandoned the notion of race as a biologically meaningful concept. Several editors supported the use of this map to make the point about the White race, and in some cases i think they genuinely misunderstood the scientific literature, in other cases they were explicitly arguing against scientific interpretations published in peer-reviewed journal articles.
So that is one example. Slrubenstein | Talk 02:42, 27 September 2007 (UTC)
I see no mention of blue eyes here: [3]
I fail to see how it is "original research" to put forward the notion that Europeans are white people nor "original research" to include material that illustrates what "white people" means in the article on "white people." This is a very good example, but not as one that supports source typing distinction. --Minasbeede 14:34, 4 October 2007 (UTC)
As compelling as that example is, I think you could make it even more obvious with a more outrageous version of OR. An article might state that the incidence of melanoma will follow this same pattern as the eye color map, without including the necessary intermediate citation "building blocks" and making some assumptions.--Filll 03:23, 27 September 2007 (UTC)
The question is, how does defining it as a primary source help communicate to someone, perhaps not overly familiar with the NOR policy, help us communicate to that user that it is an outrageous violation of policy? We are not after great examples of OR, we are after an example of where it is explicitly helpful to think in these terms. We can say, "You need to find something reliable that has made that point before you include it in Wikipedia." Do we need to say, "Do you understand our concept of source typing? If you do not, please read about it. When you have understood the concept, you will see that you have used a primary source, and in this case you have used it in an inappropriate way as you are making a point that has not been made outside Wikipedia."
I will answer that in part. The issue to me becomes that we need to think in summary style, not in discrete, citable facts. Any discussion needs context and to take facts or analysis out of context is the problem. Source typing becomes useful when we can understand the context within which an argument sits. NOR is not expressed in these terms, it thinks about discrete points. As experienced editors, we understand we cannot work in discrete elements, but need to consider the article as a whole and the source as a whole and consider each in all aspects of policy. Spenny 08:22, 27 September 2007 (UTC)
That's a useful point, and I'm certainly interested in seeing NOR framed in terms of summary style taking sources as a whole, being conscious of sourcing fact and sourcing opinions, interpretations, definitions, and arguments appropriately. The "#What is excluded?" section in recent versions[4] has been removed, and while its historical intro is replaced by the "Origins of this policy" section, the brief list under "An edit counts as original research if it does any of the following" made a useful and clear start to the article. It could be reinserted, or a better section could draw on Blueboar's ideas together with the above point. .. dave souza, talk 09:07, 27 September 2007 (UTC)
I think I might even go as far as saying that it is not the expression of policy itself that is the issue at the moment, but it is the way that the debate is distorted - people come here when a controversial issue is in dispute. I will even give you a culprit for leading the policy down a point by point path - the {{fact}} tag. This, unwittingly, encourages the examination of an individual point, rather than allowing us to stand back and say, as a whole, is the thrust of the article a fair representation of the sources. (There's more to be said, but I do witter on, so I will stop). Spenny 14:02, 27 September 2007 (UTC)
I hate to jump in here because I haven't had the chance to read the voluminous discussions here and on the sub-pages, so forgive me if I've got the wrong end of the stick...but part of the debate seems to be centred on whether the terms "primary, secondary and tertiary sources" are esoteric, or even "wiki specific". I can only say...they aren't. These are terms that every schoolchild (or at least every one who's paid any attention, i.e. about 5% ;-) will be familiar with from history and historiography courses. The meaning seems quite plain to me and these are widely used, standard terms- "factual" in particular is not at all a good substitute for "primary", if that is a proposed change. Badgerpatrol 02:52, 27 September 2007 (UTC)
Yes, they will be familiar with the terms, but not familiar with the Wikipedia definitions which ofte vary from the common definitions in each field. That's the problem. PSTS as written doesn't just ask people to learn new definitions, it asks them to unlearn old ones. Jacob Haller 03:01, 27 September 2007 (UTC)
Could you direct me to the relevant threads where examples of these departures are given? The language could be tightened up a little bit maybe, but the definitions as currently given on WP:NOR are pretty much exactly in accordance with accepted terms in historiography and information science (at least as far as I'm aware- I'm not a professional historian). Badgerpatrol 03:12, 27 September 2007 (UTC)

I also am not a journalist or a historian, but I would love to know how our definitions of primary secondary and tertiary sources differ from the standard definitions.--Filll 03:23, 27 September 2007 (UTC)


(edit conflict - a different kind of reply to JH) As I have expressed, I believe there is always room for improvement, and I have tried to suggest some ways the definitions of PST sources could be improved - and would welcome other suggestions. However, I believe it is essential for this discussion that we all agree that it is inevitable that some Wikipedia policy terms will be ideosyncratic to Wikipedia and thus require people to (temporarily) "unlearn" (I would say "bracket") their more familiar definitions. I make this as a general point and acknowledge that one can in good faith say "Yes, but this is not one of those times." We can disagree over whether we need to use the words primary and secondary, and we can disagree over how to use them. But I do not think we will get anywhere unless people asknowledge that unique definitions are sometimes unavoidable and good. I have two reasons, and an example. Reason 1: our policy must apply to people researching within or about different disciplines, professions, and domains. Thee is no reason to think that lawyers, engineers, historians and anthropologists would necessarily all use the word "source" (or pick any other important term) the same way - but we need to define it in a way that can apply to any article; in other words, we need a definition that is appropriate not for law, engineering, history or anthropology, we need a definition that i appropriate for an encyclopedia. Reason 2: we are a wikipedia, an necyclopedia written through a collaboration by diverse amateurs. Our policies must foster effective collaboration. Being diverse amateurs creates challenges and needs not faced, say, in peer-revidewed journals or EB where most authors are all academics and share many conventions, and furthermore what they write is policed by an editorial board or editors (I mean, journal editors, people with power to dictate content to authors). See - even when talking about peer-reviewed journals, we see that we define "editor" in a way utterly unlike most other people. This is already an example, but not the one i intended (which will follow) but eveytime I contribute to Wikipedia I need to "unlearn" what "editor" "really means" (meaning, the definition I need to live by in my work, if I am to kep my job and advance in my career - I have to unlearn that definition). Anyway, my point is that absent the kind of editorial supervision of journals and other encyclopedias, we rely on policies - policies take the place of people. So we are going to have policies that may be unnecessary at journals and other encyclopedias, and our policies have to accomplish something that is accomplished through very different means in other context. So it shouldn't surprise us if some of our policies are unique and use words ideosynratically. Now my real example (though editor is still a good one!): NPOV. most of us are used to this policy and realy "get it" but read the policy and you will see it is written for people who need to redefine what they mean by "neutraility" to understand the policy. So I have nothing against unique definitions, as long as they are clear and consistent and useful. I realize others may disagree with me that "primary sources" should be one of those words, but I hope we can agree on the general principle. Slrubenstein | Talk 03:29, 27 September 2007 (UTC)
The principle of primacy in learning holds that it is difficult to "unlearn" something -- it's why first impressions are so important. So if we don't need to force people to unlearn we shouldn't. The problem with the term "primary source," and the principle of primacy is that people have learned different meanings in different contexts, and they are not all applicable in the Wikipedia context. So introducing it into NOR is creates unnecessary obfuscation and confusion. I don't think your example of editor is the analogous to this, because an editor is someone who edits, and so it is a proper descriptive term in each case. Editors may edit by different rules in different settings, but editors are still editing. If NPOV is redefining "neutrality" that is also a problem (I'm not saying it is or isn't) that should not be repeated. Dhaluza 09:30, 27 September 2007 (UTC)

I have had some problems with users reading scientific papers, then taking information from these papers that occur in the "results" section of the paper. Now as far as I can see the "results" of scientific research are primary sources according to the Wikipedia definition. These results are often highly dependent upon context. The important sections in scientific papers are the introduction and discussion sections, the introduction section usually outlines the purpose of the aper, and the discussion section interprets the results. So effectively the discussion/conclusion section of a scientific paper represent secondary sources, an interpretation of the results. I have come across several occasions where editors with a pov to push take data right out of a results section, without any attempt to place these data into context, and put the data straight into Wikipedia articles, mainly because the data they use appear to support their point of view. This is a clear use of primary sources to push a pov by taking the data out of context and not having the data interpreted by a reliable scientific source. In this context the definition of what constitutes a primary source seems reasonable and pretty clear. Primary sources are documents or people very close to the situation being written about. Primary sources that have been published by a reliable source may be used in Wikipedia, but only with care, because it is easy to misuse them. For that reason, anyone—without specialist knowledge—who reads the primary source should be able to verify that the Wikipedia passage agrees with the primary source. Any interpretation of primary source material requires a secondary source. Examples of primary sources include archeological artifacts; photographs; United Nations Security Council resolutions; 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. and secondary sources Secondary sources draw on primary 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 secondary 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 secondary source. (both are from Wikipedia:Primary sources). I haven't been following this debate at all, but I was asked to comment. As far as I can see it's clear what a primary source is, and it's clear why we need to be careful when using them, so I can't see what the problem is really. Alun 03:34, 27 September 2007 (UTC)

I'm a bit confused here. Neither the term "Primary source", nor its definition, are unique to Wikipedia. It's a consistently framed, general term in common usage (within the context of history/information science). Where are these Wiki-centric neontologies? Badgerpatrol 03:37, 27 September 2007 (UTC)
I agree entirely with Alun, and having read this thread above [5] was going to say the same thing myself. Some sources may be primary, secondary and tertiary, and papers, which are highly structured to a consistent format and composed of discrete parts, are a good example. Results sections are obviously primary sources- new, first-hand research. Discussions are (basically) analysis and interpretation of that work- a secondary source. One could also argue that "Introduction" sections, which tend to synthesise previous work and lay out the aims of the study, are a kind of tertiary source. It's the abstract, discussion and conclusions that should be most relevant to Wikipedia. Badgerpatrol 03:46, 27 September 2007 (UTC)

Well, most ancient histories rely on older sources and pull them together. Some, such as Polybius or Ammianus, include some of their own observations as well as their research, but very few parts meet the PSTS definition of "primary source," or resemble the PSTS examples of "primary sources," and these mostly match the PSTS definition of "secondary source," but these are generally considered "primary sources," in history and historical archaeology. I've considered adding a note regarding ancient sources - that they should not always be considered primary sources for Wikipedia purposes - but these examples really belong in the sandbox.

Another problem is if the source is written by a participant and is published in an encyclopedia. I've sometimes disparaged tertiary sources but respect these primary/tertiary hybrids.

Another problem is when the primary source makes its own clearly-described interpretive statements. e.g. major theorists describe their own positions in their own words. Jacob Haller 06:20, 27 September 2007 (UTC)

I don't think interpretation can be considered a primary source. Look at the definitions above, primary sources more or less equate to "raw data", while secondary sources equate to "interpretations of the data". Therefore when an academic proposes a theory based on observation of primary source data, this theory represent a secondary source. Just like my description of the difference between a "results section" (primary source, ie data) and a "discussion section" (secondary source, ie interpretation of the data). I understand that in history a primary source can simply refer to the oldest extant source for a piece of information, but this is still a primary source by Wikipedia criteria because it is this source (data) that is the original for all subsequent theories. It is the theories (secondary sources) regarding what the primary source (data) means that we are interested in. If some historians claim that our primary source contains information from now non-existent sources, then it does not mean that this source is not our primary source, but it does mean that we can cite these historians as making this claim, though these claims can only be regarded as theories. We can reproduce the text of Magna Carta for example, but it means precious little without a discussion of the history surrounding it, and the text itself would be more appropriate over at Wikisource. But the text of Magna Carta is not our source, it is the source for the historians that study it, and it is their conclusions that we are interested in. Alun 06:44, 27 September 2007 (UTC)
Well, if I understand correctly, in the WP definition, primary sources group raw data with first interpretations of the data by the observer, and only reinterpretations are secondary. This is a problem because the raw data and the interpretation of it are different, but are lumped together in this definition. This is one of the key problems with the PSTS definitions, and I believe it is related to the different uses by scientists and historians. Dhaluza 09:39, 27 September 2007 (UTC)
I agree, and I'm not sure how this should be addressed. There is clearly a difference between the discussion section of a scientific paper and the primary data set. Indeed the discussion section of any scientific paper usually tries to interpret the data according to current theories, debating whether the data support or undermine any given theories, and possibly modifying any theories appropriately if necessary. So clearly the discussions sections of scirntific papers are more than just primary sources because they rely on interpreting how data fit into previously constructed models or theories. Clearly then a discussion/conclusions section of a scientific paper is not a primary source, but the Wikipedia policy may well be viewed as defining it as such. Alun 11:48, 27 September 2007 (UTC)
In the first case, the same principle applies as described above for scientific works. Those passages of the work that are the author's own observations (e.g. descriptions of ancient buildings, events that he/she has actually witnessed) or at least purport to be, are primary sources. The rest is secondary. I don't see any reason why parts of a work cannot be a primary source whilst other parts are secondary, but those with a superior knowledge of historiography than me should have an input here. The second case to me seems very unlikely- a primary source is a first hand "I was there, I did it..." account. I find it very unlikely that the scenario you envisage would actually happen (i.e. that such an account would wind up in any encyclopaedia except for Wikipedia, where it should be rooted out), and I strongly suspect that such cases are very rare- can you name some examples for discussion here if possible? Encyclopaedias are by definition always tertiary sources. The third example is not a primary source, it's a secondary one, involving as it does analysis and interpretation- even of one's own work. "We did this..." is primary. "This is what it means..." is secondary, interpretative. The Double Helix is (more or less) a primary source; the (short!) discussion section from Molecular Structure of Nucleic Acids is a secondary one, derived from empirical (= primary) results. I agree that these things can sometimes appear conceptually difficult, which is why we have pages like WP:NOR and forums like WP:RS/N to help clarify matters when things are ambiguous. I don't agree that any move away from the accepted, recognised nomenclature of primary, sec, and tert is going to help; in fact, I fear it would actually only obfuscate matters further. We don't need to be making up new words and new concepts here on Wikipedia unless we absolutely have to, and this is not one of those cases. Badgerpatrol 06:50, 27 September 2007 (UTC)
I have an old, well-used, secondhand Encyclopedia Brittanica in the other room. The article on anarchism is by Kropotkin. Many E.B. articles were written by people involved in the subject.
I find that most of what people call primary sources for political and religious theories include interpretation. In politics, if Common Sense, the Communist Manifesto, the Platform, or other works are "secondary" sources on the appropriate movements, what on earth are "primary" sources on the same?
I know that many people consider interpretive parts primary, as well as data parts. People will carry their own definitions, without checking in detail, before encountering Wikipedia definitions of these terms. Should PSTS start with the big bold passage: WARNING: Wikipedia definitions of primary, secondary and tertiary sources may conflict with other definitions of primary and secondary sources. Please read these with care. Please assess each section of each work independently of the others. Note that in politics, religion, and similar topics, there may be secondary sources without primary sources. You have been warned. ? Jacob Haller 17:19, 27 September 2007 (UTC)
Okay, that was somewhat over-the-top. But the more we rely on PSTS, the more problems different interpretations create, and the more we use it for NOR, the more our NOR-driven definitions of "primary" and "secondary" diverge from other definitions of "primary" and "secondary." Jacob Haller 17:25, 27 September 2007 (UTC)
These comments just above, IMO, are indicative of yet another set of reasons why "primary, secondary and tertiary sources" (the section on PSTS) should be left firmly in place, in keeping with both the current and previous policy. To Badgerpatrol and Alun: Are there any clearly discernible reasons why PSTS should go back to "primary and secondary sources" (call it "PSS" if we wish, maybe with derivative secondary sources or some other sub-classification of "secondary")? Currently, tertiary includes encylopedias and other compendia, and I've been very recently advocating on this page that introductory textbooks should be included in "tertiary" as well. Perhaps there are other categories of publications that belong in "tertiary"; or perhaps "tertiary" should be consolidated again into "secondary" as was the case prior to October 2006. Any thoughts on this? ... Kenosis 07:05, 27 September 2007 (UTC)
I'd include information from a scientific "review article" as a tertiary source.[6] (There's no Wikipedia article about scientific review articles oddly enough, given their importance). This sort of review article is very useful in science, but they are not intended to introduce new theories or to interpret primary sources, their purpose is to sum up the current state of knowledge in a given field of science. One major advantage of this sort of article for Wikipedia is that it will be written by an expert, unlike say an encyclopaedia article. Alun 11:48, 27 September 2007 (UTC)
Badgerpatrol, if I read yur post above correctly, you wanted an example. This may be one, Korean Wall. It's a very bad article, based almost entirely from propaganda related to tourists by North Korean tour guides. I'm not sure if this is the kind of example you asked for though. wbfergus Talk 10:38, 27 September 2007 (UTC)
To Kenosis: with a caveat stemming from my late arrival to this debate and therefore incomplete understanding of its nuances...yes, and yes. I would leave the information where it is and mainatin the distinction between secondary and tertiary sources. Badgerpatrol 11:17, 27 September 2007 (UTC)
To Wb:...and in fairness, maybe mixed with a little bit of pure OR [7] too...but the solution to bad articles with bad sources is to find better sources and write better articles. If this is am important fortification, then I would be flabbergasted if there are not suitable scholarly secondary sources (that are either balanced in and off themselves, or are at least written by pro-western authors and military historians that can be quoted as rebuttal to the North Korean claims) that can be quoted when talking about it. In fact, one might construct an argument to the effect that if such sources can't be found, then the notability of the subject is questionable and it should be deleted anyway, ipso facto. But I was specifically asking for a) examples of how Wikipedia's definitons of P, S, and T sources depart from conventional extra-Wiki definitions; b) examples of instances of primary sources ending up as articles in a reliable encyclopaedia. Badgerpatrol 10:56, 27 September 2007 (UTC)
Yeah, that article is just really bad. I ran across it way before I got all wrapped up in the NOR discussions or even knew what NOR was, so those edits of mine on the picture text are clearly OR, as is the article itself. It would be much better if the article itself was RfD'd, but I've done one or know what the proper procedure is. My edits to the picture text were to clarify what the picture is clearly about, it's not a wall, it must be a South Korean post if it's a picture of the supposed wall from North Korea, etc. but like the article itself, there is nothing published in any reliable source. Everything is strictly from 'primary sources', and nothing is cited. My further comment on the article about a vehicle falling is (what I thought at the time) merely a rephrasing of previous wording in the article, where the height of the wall is stated and that it's to serve as a bridgehead for invasion. Having been there for two years, I can say that parts of this 'structure' do exist, but not as presented in the article, but again, it's strictly from a 'primary' (or maybe more accurately), an 'original' source. In places, where the wall is along the northern side of a hill or mountain, the South Korean Army basically made a cliff along the DMZ boundary, so infiltrations or other attempts to penetrate the 'wall' would have a minimal chance of succes (read that as if attacked). The 'wall' in these areas are pock-marked with machinegun emplacements, tunnels and bunkers, etc. to prevent invasion, not to launch an invasion, but again, this is not documented anywhere except maybe in classified military files of either South Korea or the UN Command. wbfergus Talk 12:37, 27 September 2007 (UTC)

The yes, but syndrome

here in a sandbox I wrote down a summary of the problem as I had seen it. As you may see it is a list of confusions about sourcing. They may be invalid if we know what we are doing, but they are questions that have been raised that should be addressed.

I'm a simple soul at heart, and I like things to be simple. The above positions supporting the sources are all very reasonable but they do not address the problem. I get the impression that you are all used to working in an academic field and source typing is natural and obvious and it builds on your own experience of a sensible way to work.

My premise is that policy is not aimed at you. Nice though source typing is, I would say that you would all understand how to approach NOR without it. You personally would not find it essential to have the typology of sources spelt out for you: though you personally might use the technique in an informal way to test your own judgement of whether a source is safe. My analogy from the world of computing databases is that we learn about third normal form and implicitly understand it, but I would never formally use normalisation as it is obvious to me by observation.

So, if we set aside our knowledge, we look to see how the policy looks to the less informed contributor. It is here we find the "yes, but" syndrome. As I said elsewhere, a source is primary, yes, but what if it is both? Primary is an indicator of OR, yes, but not always. That fact is from a secondary source, yes, but it was posited as an original thought. It is from a secondary source, yes, but USA Today is not edited with the formal rigour of a scientific paper.

With my database example, I would not go around criticising a database because elements of it were in second normal form. In the same way that if it was in third normal form I would not accept it was an ideal design for the solution, I would want to account for the actual usage of the design, taking into account the technical constraints before coming to a judgement. I would mentally be aware of the principle, but I would not express the issue in terms of normal forms, but in terms of redundancy.

Just to continue the analogy a little further, I would also not use the rationalisation approach out of the book, because I know it leads to the most common form of error, failing to account for the time dimension of data, so people fail to recognise that the product value of a sales order item needs to be copied from the product table as the product value might change over time but the order represents a contract at a point in time. The tool is a useful guide, but it needs intelligence to apply it.

So my question to you is, yes, you are entirely comfortable with the concepts, and you see no intellectual flaw with the approach. Can you put yourself in the mind of a right-minded individual editing Britney Spears, a notable article of pop culture, and say that this policy will be obvious and useful to that editor, who may be, sad though it is, the world's most knowledgeable person on Britney Spears and simply needs guidance on why his valuable contributions are deemed unworthy? Spenny 07:52, 27 September 2007 (UTC)

I think part of the problem may be that people who are comfortable with the primary/secondary distinctions may feel that removing these definitions from NOR is somehow a judgment that this concept is "wrong" which may result in unnecessary cognitive dissonance. The point I think Spenny is making above is that the primary/secondary distinction may be applicable in academic fields (although they may be applied differently) but WP is not strictly an academic endeavor--it is an encyclopedia that anyone can edit. Yes, we may wish it to represent a scholarly work, but we will not achieve that if if the policies can only be understood by academics. So the proposals that attempt to define NOR without using the primary/secondary terminology are not meant to prove it wrong, just to show it is unnecessary in this context. It may be useful in other contexts, but the costs in terms of confusion are not justified here. Dhaluza 10:05, 27 September 2007 (UTC)
I think that once again we are nearing what many of see as the problem area. NOR should be able to define itself clearly without having additional definitions of primary, etc included in the article. These are already defined in their respective articles, as linked from the policy already, and the articles themselves (at least primary, and I think secondary as well), are also older than the NOR policy. It doesn't seem unreasonable to expect that the NOR policy should be able to define itself clearly without confusing the issue by defining (or redefining) 'subjects' that are already pretty well defined elsewhere, regardless of whether they are a 'policy' or not. Moving beyond that stage, it becomes simply "is what is being expressed in the article 'raw, or first hand' data, or is an analysis or other interpretation of that data, and in which context is this being presented". More or less in short, why do we need a 'policy', whether this page or not, to define what primary, secondary, etc. are, when there are already better pages on Wikipedia to show those distinctions with examples? Is there really a need for policy to define these, or something else more appropriate? This area has been problemmatic for a long time (almost three years at least), and wouldn't moving this one section somewhere help alleviate a lot of the various edit wars that have occured on this page, and therefore make the policy itself much more stable? Another point that been brought up is how some wording (not sure if it's in this version of the policy) insinuated that data from secondary sources should be used over primary sources. With some of the confusion about what's primary or secondary, when the 'source' can be classified as either or both, seems to be the biggest area of confusion. If a source is both, does that negate the validity of including the work from that source over another source? It's kind of a "this source trumps that source" argument. If an article on aircraft includes things like engine thrust, is it preferred to have the engine thrust data from the article/journal or instead have it come from the manufacturers spec sheet (especially if there is a variance in the data)? The journal article is a secondary source that repeats the primary source (maybe with a typo), so is that still preferred over the original data? Well, time for me to head off to work. I might be able to continue from there depending on how busy I get. wbfergus Talk 11:04, 27 September 2007 (UTC)
I can't speak for anyone else, but I was taught about primary, secondary and tertiary sources in (compulsory) history lessons when I was about 13 (and believe me, my schooling and education were unexceptional in oh so many ways...;-). I very strongly suspect that the concepts are taught to schoolchildren the world over (let's not forget, judging and classifying sources is fundamental to the study of history and is one of the most important transferable things they teach you). This is not some specialised, esoteric academic doublespeak. I would bet that the vast majority of people reading the material already have at least the kernel of an idea regarding these concepts, and as currently written it seems fairly straightforward. There will always be an element of good judgement involved, but we can't legislate for individuals' personal judgement. We should perhaps stress more clearly that user and article talk pages, as well as forums like WP:RS/N are available, and should be used for constructive debate regarding the nature of a source and whether it should be used. Even if our savantic friend User:Britneysthebest178 doesn't understand how to classify his sources (and there's no shame in that), there are plenty of others that do and can help him or her. Badgerpatrol 11:17, 27 September 2007 (UTC)
I think that is common ground, and though I would strongly disagree on the small point of classification of sources being, let us say common knowledge, it was simply not a concept put forward when I sat for two years having the Napoleonic Wars, Nelson and the Agricultural and Industrial Revolution dictated to me. At least I knew Jethro Tull wasn't a band.
Whilst I am not convinced, the consensus is that source typing is (at least) a useful tool to assist. I think we can say there is consensus that it should be properly associated with the various different guidance in some way.
BadgerPatrol, though it is useful, would you say it was necessary? And if so, why? Spenny 11:45, 27 September 2007 (UTC)
Maybe I touched on it briefly in a previous post above, but could the NOR policy itself more clearly and accurately describe itself by using something like "original data" vs. "analyzed data (conclusions, etc.), preferably peer-reviewed when available" instead of delving into re-defining primary, etc. when such definitions already exist in Wikipedia are are linked already? Add a statement like "For further definitions see....". I think this is more along the lines of what many of us are trying to say, though this idea hasn't yet been clearly and succintly expressed (if it can be). I could be wrong though, I'll let others comment on if maybe this might be what some of us are trying to work towards, without weakening the policy. I also think that many of us from the groups 1 & 2 talked about previously, still fail somehow to see how moving the PSTS section from NOR (to some other, more appropriate place, not a deletion), would weaken the NOR policy. The policy itself is purely about not allowing original research into a Wikipedia article, not about how to classify different sorts of 'source material'. While one may help the other and be otherwise related, they are completely different subjects, like having an article on the Korean War with a distinct sub-section included in it that goes into detail on what weapons are, what the different kinds of weapons are used for and how they used, etc. It may be usefull knowledge, but really has no bearing on the article (or in this case, policy). wbfergus Talk 12:50, 27 September 2007 (UTC)
Yes, but...I was not taught this; or if I was, I certainly didn't retain it. And I was fortunate enough to attend good schools, so that's not the reason. So I tried looking it up, and the first hit I found had this definition:
What is a primary source? A primary source is a source that was created during or immediately after the event or period it documents.... Primary sources can be distinguised from secondary sources, which are are sources created at some later time. A description of Ellis Island at the beginning of the 20th century, if written today, would be considered a secondary source (although it might be based on primary sources.) Primary sources are useful because they can give detailed information about a place, time period or event, as well as because they give us insight into the views and experiences of people without showing them through the lens of later events.
So this discusses primary/secondary strictly in terms of time, and has nothing to do with interpretation. And it actually favors primary sources in this context. I don't think the definitions are as widely known or as unambiguous as some assert. Dhaluza 01:46, 28 September 2007 (UTC)
First 10 links when googling "primary sources definition": [8] , [9], [10], [11], [12], [13], [14], [15], [16], [17]. Those look pretty consistent to me. Once could make a legitimate argument against distinguishing between the concepts of secondary and tertiary sources, which is not always done, although I would personally retain the distinction. But the concepts of primary and secondary sources are crystal clear and categorical. These are commonly used terms in historiography. These links [18], [19], suggest that the discrimination of primary and secondary sources is taught to children (in the UK) as part of the national curriculum at or about key stage 2 level (7-11 years old). Any history teachers lurking would be able to comment on this with more professional knowledge than I have however. I struggle to accept however that these concepts are not therefore taught at a similarly early age to children studying history elsewhere in the world. Badgerpatrol 02:58, 28 September 2007 (UTC)
Yes, but I just don't recall it being taught in those terms, but that was also a long time ago, and I don't remember a lot of things from my school days. I was using these Google search terms to try to see if it was in the current curriculum (without success) when I found the ref above. It was actually the second hit (but the first I clicked through); looking back at the first I found: "Primary sources were either created during the time period being studied, or were created at a later date by a participant in the events being studied (as in the case of memoirs) and they reflect the individual viewpoint of a participant or observer." This is a conflicting definition because memoirs are classified as primary not secondary based on proximity rather than time. It also says that primary sources contain interpretation, which needs to be treated differently from facts from a NOR perspective. So although this definition is similar to the PSTS definition here, I think it points out some of the problems with using PSTS to describe NOR.
Looking at the first ref you give, I found this: "Primary resources provide firsthand evidence of historical events. They are generally unpublished materials such as manuscripts, photographs, maps, artifacts, audio and video recordings, oral histories, postcards, and posters...." This definition of primary sources is not helpful because it defines primary sources primarily as unpublished materials, which are not allowed as WP sources because they are unpublished, not because they are categorized as primary. Your second hit says: "Material from, or directly related to, the past...." which again is not helpful, because we also cover current topics, such as aircraft still in production. Here factual statements like manufacturers' performance data (e.g. range (aircraft)) should be treated as primary, even though it is relevant in the present, not just the past. I could go on and on, but I think the point is made that the definitions are not nearly as cut and dried as you assert. And these differences lead to irrelevant arguments over primary vs. secondary, rather than the relevant issue of OR vs. NOR. Dhaluza 10:49, 28 September 2007 (UTC)
No, I don't see anything in any of those definitions that differs in essence from the definitions we currently have. Time has nothing inherently to do with whether a source is primary or secondary, although it is true that many primary sources are composed at or near the time of an event. All that matters is that it be a first hand account, i.e. by a participant. Memoirs written years after an event are still a primary source. This doesn't jar with our definition. It's not claiming that primary sources contain interpretation as such, its rather claiming that they are (in the vast majority of cases) biased. Archaeologists select the most attractive and valuable objects from sites, not a random spread (usually) anyway. Photographers take the most striking photographs, not a random spread. Politicians tend to maximise their reputation when constructing their memoirs. Unlike the concept of P and S sources, the concept of what is a "fact" is very difficult and really would lead us down a blind alley. As a a great philosopher probably never said, one man's "fact" is another man's total bollocks.... That definition is merely stating the obvious- that just because something is included in a primary source, it doesn't mean it's a "true", unbiased account. That doesn't jar with anything we currently have, although the shortcomings of primary sources could be expanded upon. As for publication- the definition you quote specifically does not demand that primary sources be unpublished. Generally they will be unpublished (i.e. not everyone can see them and make a judgement), which is one of the many reasons that the use of primary sources is discouraged on Wikipedia. Finally, the time issue. By definition, all sources were written in the past, whether they be primary, secondary or tertiary. If the are not from the past then they haven't been written yet. Except that it rather obtusely states the obvious, I don't see how that jars with either common sense or our current Wikipedia definitions. Badgerpatrol 11:37, 28 September 2007 (UTC)
Archaeologists don't do that, though some used to. Selecting the most attractive/most valuable artifacts throws off future studies of the proportions and distributions of artifacts. (For example, how common are North African amphorae - or fragments thereof - compared to Italian amphorae in Spain?) However, the artifacts are no more important than the excavation notes showing their context. Jacob Haller 17:44, 28 September 2007 (UTC)
I would beg to differ with "Time has nothing inherently to do with whether a source is primary or secondary, although it is true that many primary sources are composed at or near the time of an event." According to Internet Public Library (New York City History) (see text and link on Wikipedia:No original research/Sandbox/Various examples, time can make a distinction between primary or secondary. It all depends I guess on which school of though you 'learned'. I never did until I first came to this policy a tad over a month ago, this is all new to me, and I'm almost 50. wbfergus Talk 12:06, 28 September 2007 (UTC)
That's just their bad phraseology- their idea of "immediate" can be taken to mean "within the author's lifetime". They are not suggesting that memoirs written by a participant 50 years after the event automatically become a secondary source. They are saying that biographies of one's life written by one's grandson 50 years after the event are a secondary source. Their school of thought is the same as everyone else's. Badgerpatrol 13:21, 28 September 2007 (UTC)
I read it as being started by students, but now it is more in line with "peer-review', being a collaborative project of numerous Universities ([22]), but that's besides the point. It does show though that there are differing schools of thought. These are some of the people who may very well go to write some paper or whatever using these terms and definitions, and in a couple years we'll be faced with additional verifiable sources that state something different (which I think is what got us here in the first place). Why can't OR (or NOR), be clearly stated without delving into the contentious realm of what is a primary source for which discipline vs. what is a seconadry source for that same discipline? That really seems to be the crux of the problem. One school holds that OR can't be defined without resorting to broad statements about the type of source, instead (from the other school) of just making a simple case about the proper use of the source. wbfergus Talk 13:46, 28 September 2007 (UTC)
It was in fact written by a student at the University of Michigan from what I can gather, but I agree that she should know better. That website is not even remotely close to peer-reviewed, but I see your point. All I can say is...it's not contentious. The definitions we have now are perfectly accurate and are universally accepted. The fact that there are some incorrect definitions available to be found on the internet suggests more about the internet (i.e. a significant amount of it is nonsense) than it does about the matter at hand. The reason to include a discussion of source types in WP:NOR is simply because anybody can potentially select and juxtapose primary sources (which may not be inherently unreliable per WP:RS) without context or analysis in order to advance their point of view. That's OR. Badgerpatrol 14:21, 28 September 2007 (UTC)
First, a student can also contribute to WP, and they will bring their own definition of PS with them. To imply that she is wrong and you are right, and then dismiss it saying the definitions are universally accepted is self contradictory. Second, and more importantly, in your OR example, the problem is selecting sources to advance a point, whether primary or secondary. The source type has nothing to do with the OR (except that some believe it is easier or more likely with primary sources, but this is not universal either). There is nothing wrong with juxtaposing a representative sample of related facts in an encyclopedic context and letting the reader form their own conclusion. Dhaluza 00:24, 29 September 2007 (UTC)
Forgive me for being somewhat dense this morning (for me anyway). Which definitions are you referring to, that we have now? The definitions of what OR is, or the definitions of what type a source is? wbfergus Talk 14:42, 28 September 2007 (UTC)

<unindent>The definitions of primary, secondary (and to a lesser extent) tertiary sources as currently written on WP:NOR. The page in general could do with some clean-up- in particular I dislike any policy that makes reference to deus ex Jimboina quotes as if unquestionable wisdom; there are better ways to present these points. But the definitions of P, S and T are accurate and useful, in my view. Badgerpatrol 14:56, 28 September 2007 (UTC)

Ah, okay. Then I would tend to disagree then that they are "perfectly accurate and are universally accepted". The Lafayette College Libraries and Academic Information Resources page has some good information, and examples, of how the term is used differently among different discplines[23]. I think that additionally, the section on the other page they have a link to (Evaluating sources), which has a section called "How does the source under study differ from the original?". This starts off with "It is always important to consider how a primary source surrogate differs from the original. In some cases, this may have a significant effect on how an item can be interpreted. Primary sources exist in too many different forms to present anything but a suggestive list of issues to consider when evaluating how well a surrogate represents an original source". This (I think) helps show that Wikipedia's definitions (specifically the re-definitions within this policy), while being close, are not "perfectly accurate" and if there are so many variances in the definitions, then how can Wikipedia's be "universally accepted" as well? wbfergus Talk 15:53, 28 September 2007 (UTC)

PSTS and conflicting definitions

Some people have questioned why we do not simply use "the standard definitions", or how the wiki definitions differ from academia. The problem is there is no standard definition for the distinctions in academia, as it is highly relative to the field. In the natural sciences, peer-reviewed articles (outside of review articles) are usually considered primary sources. In the general humanities, peer-reviewed articles are considered secondary sources. In historiography and some branches of literature, peer-reviewed articles are considered secondary sources. In information/library science and some branches of literature, peer-reviewed articles are primary or secondary based on whether or not they make novel claims and/or present new data. This is only a very rough overview of the distinctions and does not come close to encompassing the various conflicting definitions used for the terms throughout academia.

These conflicting definitions can be problematic. A scientist could very well have a strong amateur interest in history and decide to edit some of those articles. Due to their background, they have learned that peer-reviewed articles presenting novel research are primary sources. However, a historian working on the same article will view peer-reviewed articles, regardless of novel assertions, as secondary sources. Let us assume they both have a solid understanding of what constitutes original research. In the absence of PSTS terms, they could both probably come to clear agreement about what is original research in the article and address it without much disagreement or even much discussion. Introduce the PSTS terms and it is easy to see how that discussion could bloat considerably into a debate about what constitutes a primary or secondary source. Please also see some of my previous comments to get a better understanding of my concerns.[24][25][26] Thanks for reading. Thoughts? Vassyana 17:28, 27 September 2007 (UTC)

That gets to the heart of my opinion on the issue. I don't particularly care what is called a secondary source and what is called a primary source, so long as the sources that have always been acceptable are still acceptable and the sources that have always been unacceptable are still unacceptable. — Carl (CBM · talk) 17:43, 27 September 2007 (UTC)
I don't want to pre-judge any outside input we may get, but I think there is a point being missed here. Whether a source is primary or secondary has nothing to do with the discipline of the author. It has to do with the nature of the source. Universal rules as to what constitutes a primary, secondary (or tertiary) source exist and can be applied regardless of discipline. History scholarly articles are always secondary sources, drawing from primary sources. One can't make history- but one can "make" science through novel interpretation and experimentation. Where one is quoting from a specific part of a scientific paper, then a distinction may be made between primary and secondary sources- e.g. the results are primary and the discussion and introduction may be secondary. Science review articles synthesise previous work and are thus secondary sources. Hopefully I'm articulating my point- the same rules apply, regardless of the discipline, because the classification is at the level of the individual source itself. Badgerpatrol 04:24, 28 September 2007 (UTC)
Carl, the issue here is not so much which sources are acceptable or unacceptable, but rather the range of inferences that can be drawn from sources without being considered original research or original synthesis. WP:NOR says in relation to use of primary souces: "An article or section of an article that relies on a primary 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 primary sources should be careful to comply with both conditions." In other words it has to do with how the sources are used so as to avoid OR. ... Kenosis 18:55, 27 September 2007 (UTC)
You know, that passage bugs me. Read literally, I think the statement means that the addition of one primary source mandates the removal of all interpretive passages from the article, and vice-versa. AFAIK, the real point is that an interpretive claim in the article requires the same interpretive claim in one of its sources. A non-synthetic claim is at least as strong as the strongest of its sources.
In addition, it's not uncommon to add unreferenced material to an article, which is a reasonable interim measure, and it is not unknown to add additional works to the references list without checking them against the article, which strikes me as incredibly irresponsible: it makes it look like the article was referenced against book and the book was used in writing the article, when neither may be true.
Because some people do this, other people, including myself, only count in-line citations, and ignore other citations. Anything which doesn't have an in-paragraph citation, and has been questioned, is an unreferenced passage.
I think most editors would agree that the unit of citation is not the article or section but the paragraph, quote, sentence, or as low as needed. Jacob Haller 03:24, 28 September 2007 (UTC)
A "unit of citation"? Interesting concept. The most I can make of this commentary just above by Jacob Haller (beginning with "You know, that passage bugs me." ) is that the words "an article or section of an article" in the above-quoted passage in WP:NOR, about primary sources, might be changed to something like "an article or part thereof" or something that makes clearer what is being referred to in that passage. I am fairly certain it is intended to refer to an "issue" or "fact" or "particular aspect of presentation in an article", or something like that. This is a minor "sofixit" issue, IMO. ... Kenosis 05:12, 28 September 2007 (UTC)
Kenosis, in the other post above you said "In other words it has to do with how the sources are used so as to avoid OR". Isn't this oversimplifying it? Isn't it more in line with "it's how the text is used" instead? Couldn't I state that Bill Clinton "never had sex with that woman" and quote from his nationally televised statement (primary source) and numerous followup articles on the statement (secondary source). If I instead said that while Bill Clinton "in fact did not have sex with Monica Lewinski, she actually performed the oral sex and Bill Clinton was just the willing recipient", that's the OR, not the sources. Is that correct? wbfergus Talk 12:52, 28 September 2007 (UTC)
Um, look, the unit of citation wasn't my main concern. As long as we're sticking to the sources, the claim is at least as strong as the strongest source, and more sources can complement each other. Adding more sources doesn't hurt. Suppose one sentence makes a factual claim and an interpretion of it. If we have one source which supports both, and we add another source which corroborates and strengthens the factual claim and can't support an interpretive claim, that does not mean we must remove the interpretive claim. Jacob Haller 17:58, 28 September 2007 (UTC)
See WP:NPOV, WP:NPOV#Undue_weight, WP:VER#Sources, and WP:CON. These kinds of editorial decisions need to be made on a case by case basis. In the example just given, the source that supports the evaluative claim removes it from the realm of original research and puts the remaining decisions in the realm of the other policies I just mentioned. ... Kenosis 18:43, 28 September 2007 (UTC)
Okay, say the sames sources were cited for both examples, the first merely repeats what the sources say, but the in the second example I'd be rephrasing it incorrectly, which gets to the heart of the OR matter. It's not the source per se, but how the source is used, isn't it? I'm not trying to argumentative (though it may appear so), I'm trying to figure out exactly how source types are either inherently good or bad. I've seen this 'position' (for lack of a better word) numerous times in the last month or so, but I don't remember anybody having provided a good example to that effect. wbfergus Talk 22:57, 28 September 2007 (UTC)
Yes, but as written one primary source "spoils" the whole thing. If that's not the intent, can we come up with an alternative to fix it? Jacob Haller 18:59, 28 September 2007 (UTC)
It never was the intent of WP:NOR to mean one couldn't throw in a supplementary primary source in further support or verification of the non-analytical portion of a composite statement in an article. ... Kenosis 19:06, 28 September 2007 (UTC)
(outdent) I can see that the bit saying that articles or sections relying on primary sources mustn't make evaluative claims (etc etc) may be read to mean what Jacob is saying. It should probably be clarified, and if it weren't for the protection I'd be bold and do so - it should probably say that articles or sections relying solely on primary source mustn't do this that and the other. SamBC(talk) 20:38, 28 September 2007 (UTC)
Yes, since it's been pointed out, it should be stated so as to make this clearer. I imagine the addition of the word "only" or "solely" would do the job here. ("An article or part of an article that relies only on a primary source should ... " ) ... Kenosis 22:41, 28 September 2007 (UTC)
I assume that, in this context, you mean acceptable and unacceptable relative to usage - even in the policy as it currently stands, and over most of the life of it if my reviews of history aren't misleading, primary sources have been perfectly acceptable as long as they aren't used to generate novel conclusions (ie, to generate original research). SamBC(talk) 17:52, 27 September 2007 (UTC)
Vassyana, that's a fantastic summary and statement of the reasoning that many of us have come to understand without, it would seem, being able to articulate it well. What occurs to me about this is that it might be useful to re-frame the current section so it doesn't read as a definition of OR. I don't think it currently is a definition of OR, but experience shows that it is read as such often enough to cause confusion. This might be acheived by changing where in the article it appears, and/or adding/changing the section introduction to make it clear that the section gives guidance on the impact/role of sources to OR questions, rather than actually specifying the policy or defining OR. After all, if one considers OR defined, the specification of the policy can be framed incredibly simply, and I think that people will generally agree that it's straightforward to define what OR is without recourse to PSTS. PSTS helps clarify and explain the definition, or rather some aspects of it, rather than being part of the definition. Do people think that sounds reasonable? SamBC(talk) 17:52, 27 September 2007 (UTC)
Could we get Vassyana's summary copied over to the sand box page... I am sure others would find it helpful to refer to, and I don't want to lose it among all the other arguments and side discussions. Blueboar 18:06, 27 September 2007 (UTC)
Done: Wikipedia:No original research/Sandbox/Change needed#Vassyana's Summary (2007-09-27) SamBC(talk) 18:25, 27 September 2007 (UTC)
One draft referred to "statements of fact" and "statements of interpretation." I don't recall whether that lasted into Blueboar's proposal. But for source-typing, we need to consider:
  • Through the entire source: Reliability
  • Through most of the source: First Party or Third Party
  • Section by section: Context
  • Statement by statement: Statements of Fact or Statements of Interpretation
  • I'm not sure where PSTS proper fits in this typology. Jacob Haller 18:47, 27 September 2007 (UTC)
A lot of those points aren't actually relevant to OR - whether a source is reliable or not doesn't affect whether using it is OR. It's not acceptable from unreliable sources, but it's then not original. The same is true for first-party and third-party, generally. Context and statements are, however, important. I believe that one way of summarising the issue with primary sources is that, whatever they look like, statements from primary sources can only be statements of fact. Whether this is accurate or not, I do not comment on at that point. I'm just illustrating the explanation of why PSTS matters in terms of the actual "teeth" of the policy as it stands. SamBC(talk) 19:02, 27 September 2007 (UTC)
I posted above another conflicting definition I stumbled across. At bottom, I think this is very simple. We can only use statements of fact as statements of fact. What else is there to say? Dhaluza 01:59, 28 September 2007 (UTC)
An article or section of an article must not rely on analytic, synthetic, interpretive, explanatory, or evaluative claims from documents or people very close to the situation being written about, but can report facts about opinions without asserting the opinions themselves by attributing the opinion to someone and discussing the fact that they have this opinion. .. dave souza, talk 11:38, 28 September 2007 (UTC)
I'd agree that no part of an article should rely on analytic etc claims from sources close to the situation. I expect we all would. However, that isn't actually a matter of original research, is it? I mean, the editor isn't conducting original research by using them. The sources are inappropriate for other reasons, but it just isn't a matter of original research. SamBC(talk) 16:11, 28 September 2007 (UTC)
Sorry for being imprecise. I meant what else is there to say about factual sources. I agree that interpretation etc. also needs a source, and an independent source is best. Dhaluza 23:47, 28 September 2007 (UTC)
I added some definitions (one from this talk page, two from external sites) over at Wikipedia:No original research/Sandbox/Various examples. Not sure if this helps or not, just a few examples I decided to post after seeing once again that some say there is a difference across disciplines while others say there isn't. wbfergus Talk 11:52, 28 September 2007 (UTC)

Outside input from scientists, historians, information scientists

I'm not really convinced that these distinctions between disciplines as to how sources are viewed actually exist or are relevant. To get more input, I've left messages at on the talk pages of Wikiproject History, Wikiproject Science, and Wikiproject Librarians, inviting experts in those fields to come here and give us their input. Hopefully their input will be forthcoming and will help us get a consistent idea. I personally have never come across definitions that differ in essence from those we already have on WP:NOR. Badgerpatrol 03:24, 28 September 2007 (UTC)

Good idea. Let's wait and see what they have to offer. ≈ jossi ≈ (talk) 03:38, 28 September 2007 (UTC)
Badgerpatrol, I think you are glossing over significant differences as I outlined in more detail in the section above, explaining how these differences are relevant, or more specifically how the definitions are irrelevant to NOR. But I also think this is a good idea, and I would like to see more diverse opinions too. Dhaluza 10:44, 28 September 2007 (UTC)
While you are beavering on bolstering up the understanding of source typing, and it is useful for understanding, this path is missing two things:
  • (1) a response to the question, politely phrased as "Do you think source typing is essential for OR policy?" - or as we have suggested more informally, does it pass the "So what?" test.
  • (2) you are still drawing from a biased sample. To establish understanding of source typing, how about sampling those outside the spheres of science and history? I'm not sure how you do this, as I think it is likely to be so uninteresting to Mr Joe "I changed my name to Britney Spears by deed poll" Public that you do not get a fair sample. Even if we establish compatibility between history and science, we have not addressed a wider issue.
The suggestion that people do not understand primary and secondary has been disparaged as absurd, but the point of such shorthand terminology is that it carries a lot of baggage with it, we can recognise the cardinality (or is that ordinality?) of the suggestion, but we have to go into some detail to ensure the point of it is clear. Spenny 13:29, 28 September 2007 (UTC)
No outside input yet, so let me put forth my hypothesis: Historians are concerned with the time frame of the source--whether it was written in the time period, or later. Information scientists are concerned with proximity--whether the source is from someone close to the subject or event, or someone independent of both. Research scientists are concerned with the difference between observations and conclusions of the original experimenter, and reinterpretations of the published data. These are different perspectives of primary vs. secondary sources, and the WP PSTS definition conflates these to produce a homogenized one size fits all definition. Dhaluza 00:43, 29 September 2007 (UTC)
I work in mathematics and regard research papers in my subject as both primary and secondary sources. These papers typically interpret old results as well as proving new results; the primary/secondary distinction is not particularly valuable to us. I don't particularly care if the guideline calls them primary or secondary so long as they can still be used in the ways they always have been on WP. — Carl (CBM · talk) 00:54, 29 September 2007 (UTC)
Carl, does that mean that you wouldn't care if the PSTS section was moved out of this policy to a more appropriate (or logical) 'home' (existing or new policy, etc.), or do you have a preference to still keep the section here? In your work, does the 'type of source' classify the contents as 'original research', or does 'original research' become a criteria for what the 'type of source' it must be? Thanks. wbfergus Talk 11:35, 29 September 2007 (UTC)
I don't have any strong feelings about whether that section should stay here or move. In my area, mathematics, I can divide research papers into two rough categories:
  • Papers that are primarily intended to convey new results, but will also discuss and interpret older results.
  • Papers that are primarily intended to discuss and interpret older results, but may also contain some previously unpublished results.
These papers can be used completely interchangeably as references in other papers. On Wikipedia, we use these papers both for the statements of fact and for the interpretations they give. We decide on a case-by-case basis whether particular interpretive claims can be stated as facts, with just a citation, or must be explicitly described as the opinion of the author of the paper. — Carl (CBM · talk) 14:29, 29 September 2007 (UTC)

minor proposals

I'd like to add the following at the end of the "origins" section:

As a more diverse group of editors were drawn to Wikipedia it became clear that other topics besides physics, such as politics and religion, were attracting original research, and the community sought a more systematic way to define original research and to guide editors in avoiding it.

Given that our discussion has involved confronting different ways people in different disciplines use sources and do research, I think that this should be uncontroversial. Any objections? Slrubenstein | Talk 13:05, 28 September 2007 (UTC)

None here... I would add some topics that are not so obviously controvercial to the list (history/pseudohistory articles come to mind). Blueboar 13:16, 28 September 2007 (UTC)
I agree with you and if you feel strongly go ahead and add "history." I don't only for this reason: I tried to limit myself to just a few examples so as not to give the imporession that I am even attempting an inclusive list. My concern is if we add history people will say "Well, what about anthropology? Sociology? Journalism? Law?" and so on and so on. I just wanted to suggest a more diverse group. And I wanted to keep it simple. But if you can think of a better wording I won't object. Slrubenstein | Talk 13:22, 28 September 2007 (UTC)
I've read it and re-read, trying to determine if any hidden meanings were escaping me (I'm rather dense at times, especially before I get my caffiene levels to normal). I don't think I have any objections to it. It did get me to thinking though if perhaps it might be appropriate (since we are discussing the history of this policy), to also note that there has ensued numerous 'discussions' (or arguments) on the types of sources and their usage and/or preference, since at least January 2005 (the oldest talk page I can find). wbfergus Talk 13:31, 28 September 2007 (UTC)
No objections from me. ≈ jossi ≈ (talk) 23:26, 28 September 2007 (UTC)
It looks OK to me, except the whole thing is OR, which is ironic. Probably the whole history should be converted to summary style, and the detail should be moved to another page that can cite the diffs. Dhaluza 23:51, 28 September 2007 (UTC)
Please don't forget [this thread], where J. Wales deals with OR as it relates to history. Incidentally, the current location of the "origins" section is an appropriate spot for it, IMO. Had the participants here revisited that issue as yet? ... Kenosis 01:44, 29 September 2007 (UTC)

Since no one objects I am putting it in, along with the thread Kenosis constructively provides. I don't think there has been any discussion about placement of the section. Until there is, I will leave it where it is. Slrubenstein | Talk 09:49, 29 September 2007 (UTC)

A minor point, but the ref that's included is not in a proper format, throwing off the numbering by having two refs labeled as [1]. wbfergus Talk 11:27, 29 September 2007 (UTC)

Can Original research be defined without defining the types of sources?

I just went over Google and search for +"original research" +definition. There are couple of interesting hits (and a bunch of 'bogus' hits). One 'hit' that doesn't delve into what the 'source-type' is, is [27]. It is also interesting to note a somewhat similar discussion taking place on [Wikiversity:Scope of research/En]. There appears to be numerous examples of how other 'entities' seem to be able to define this term fairly easily without resorting to what the type of source is (or was). If other 'entities' can do so, why can't Wikipedia? wbfergus Talk 14:15, 28 September 2007 (UTC)

That link does nnt work. In any case, "Original research" is a term coined in WP to address a specific constrain in the aims of the project. ≈ jossi ≈ (talk) 23:28, 28 September 2007 (UTC)
For me, that link brings up a University of North Florida Library student guide page which is headed "What is Original Research? Original research is considered a primary source." What was that about it not delving into what the 'souce-type' is? .. dave souza, talk 07:49, 29 September 2007 (UTC)
The point was, can this policy define NOR without getting into the definitions of different types of sources (like other 'places' seem to be able to do, this is not a Wikipedia-specific term). While the first link does start with "Original research is considered a primary source", it does not state that use of other primary sources is original research. Nor does it anywhere state any definitions of primary, secondary, tertiary or anything else in those terms, though it does give a brief definition of journal, magazine, etc. without 'typing the source'. The only reference that is made to 'primary source' is that original research itself is a primary source. This goes hand in hand with WP:NOT#OR, which says that Wikipedia is not a primary source. The second link, to Wikiversity, was merely included to show that another Wikipedia (sister project?) is also contemplating NOR issues, but instead of primary source, etc., there is a reference to "original research"/"secondary research" which may be of use (I don't know). Also, while the link I first provide was the correct page header (or article title, like we use here), the actual URL is [28], which appears to be a beta version of Wikiversity. wbfergus Talk 11:21, 29 September 2007 (UTC)

The problem with PSTS—Insufficient granularity

I posted a hypothesis above that the definitions of PSTS, and specifically the primary/secondary distinction in NOR, do not distinguish between time, proximity, and interpretation, but instead conflate these into a homogenized definition. For example, a diary is written in the moment, and memoirs are written later, but both can contain observations and interpretation by sources in close proximity to the subjects. An authorized or auto-biography is similarly close interpretation, but an independent biography has more editorial distance. A research paper presents original data and analysis by that is close and recent, but a review paper presents independent analysis at a later time. A news report is written by an independent journalist at the time of the event with minimal analysis, while an editorial is also timely but primarily analytical. A historical account may be written at a much later time, with or without additional analysis.

The problem is that time, proximity, and interpretation are three somewhat independent variables, but there are only two possible output states, primary and secondary (ignoring tertiary which is a different case). If we take the inputs as binary variables, there are eight different permutations. What is needed is a taxonomy of sources, or more specifically their use, that provides more granularity than PSTS offers. Dhaluza 01:34, 29 September 2007 (UTC)

As you indicate, it's all relative to context, and setting the line is a judgement call. You seem to want to substitute a complex and rigid taxonomy for judgement and discussion to reach a consensus. However, the output is binary - whether a source can only be used for obvious fact, or whether it can be used for analysis, opinion or conclusions about the facts which are the subject of the article or section. ..dave souza, talk 07:59, 29 September 2007 (UTC)
So then in the context of NOR, are you saying we should only be concerned with whether a source is factual or interpretive, and analysis of time and proximity are not related to NOR (but may be important for other reasons)?
I was thinking that if we must discuss sourcing in detail in NOR, as many have argued here, it needs to be a honest analysis, rather than an oversimplified one like the present PSTS. That would suggest a broader taxonomy, but not necessarily a complex and rigid one. The NOR policy at present also seems redundant and repetitive, and it might be more clear under a "unifying theory" that better integrates the various points.
The conflation of time, proximity and interpretation just leads to pointless arguments over where to draw the line between primary and secondary. I think it's been fairly well established that a source may have both characteristics, and a secondary source can be used as a primary source about itself. So, although PSTS was probably intended simplify source judgments in theory, the devil is in the details, and by glossing them over, PSTS actually makes source judgments more complicated than they need to be. Perhaps a broader taxonomy would be more clear, and ultimately more concise as well. Dhaluza 08:22, 29 September 2007 (UTC)
perhaps, rather, a more specialized taxonomy. In the natural sciences, its fairly clear that the primary sources means the primary journals in which research or calculations are reported. The actual ultimate material being research is the data, not a publication. And a secondary sources is either a review article, or an indexing and abstracting service. In history, the primary sources are the data being studied, the narratives or monuments. Everything else is what is wrtitten about them , and secondary. The WP definition is an uneasy blend between the two. DGG (talk) 09:05, 29 September 2007 (UTC)
Uneasy blend or not, the WP definition is simple and explicit: "Primary sources are documents or people very close to the situation being written about." Answering Dhaluza, I'm saying that in the context of deciding if the source as being used (not necessarily the source as a whole) is factual or interpretative, we also have to decide if the source is very close to the situation being written about. If it is, we should only rely on it to make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge. Thus, if the source is the subject of the article or section and makes descriptive or analytical claims or conclusions, we must report these as the claims made by the source, and look for secondary third party sources for analysis. It appears that you know this, but want it put into another policy. However, it's directly related to the evaluation and use of sources that are dealt with by NOR. It does seem that the terms have specialised meanings in certain academic contexts, but people familiar with these contexts should be erudite enough to accept that their meaning isn't universal. A statement to that effect might strengthen the policy. .. dave souza, talk 11:49, 29 September 2007 (UTC)
OK, so I think you are saying both interpretation and proximity of the interpretation are important, which is reasonable. The case you mention is where we use a close interpretive source only as a factual source about itself. I hadn't thought that that belongs in a separate policy, but on further reflection, maybe it does. This case is actually not OR in that it is original in relation to the subject, not the WP editor, and maybe this is part of the confusion here. This really overlaps with WP:SPS and WP:SELFPUB, and perhaps the scope there simply needs to be further clarified to cover self-produced material, even if it is published by another party without independent analysis. This may simply be a case of multiple definitions of the same problem in different policies. WP:NOR should focus on WP editors putting their own opinions in articles, and WP:V can address inclusion of non-editors' opinions about themselves. Dhaluza 13:23, 29 September 2007 (UTC)
Dave... I think I see what you are trying to say... but I have a question. You say: "Thus, if the source is the subject of the article or section and makes descriptive or analytical claims or conclusions, we must report these as the claims made by the source, and look for secondary third party sources for analysis. ... it's directly related to the evaluation and use of sources that are dealt with by NOR." How? I agree with your statement that, if a source is the subject of the article, we should note that fact and attribute the claims or conclusions made to the subject, but how is this tied into NOR? As Dhaluza points out, such analysis and/or conclusion is that of the source and not the WP editor who discusses and cites to it. NOR is about the originality of statements made by Wikipedia editors... not about the originality of statements made in a source. As long as a claim or conclusion is made externally to wikipedia, discussing and citing to that claim isn't OR. It might fall under WP:Fringe, it might be unreliable under WP:V and WP:RS, but it doesn't constitute OR. Blueboar 14:17, 29 September 2007 (UTC)
Well, Blueboar, it's OR to the extent that the editor is selecting or synthesising their interpretation of the subject and not finding a third party interpretation. WP:V doesn't cover it at present if the source is reliable, published by a reliable publisher rather than being self-published in the sense set out in WP:V. Thus for example a biography of the subject can provide analysis, comment and conclusions, an autobiography can only provide facts however worthy the publisher. Similarly, in writing an article about, say, a science book which itself is a secondary source on its subject, we can report obvious facts from the book, but must look to a source writing about the book for secondary analysis. It ties in with the evaluation of whether we're reporting facts or reporting evaluation etc. which your proposals set out, and in my opinion fits well with them as an expansion on the description of OR which they could provide as a first section after the intro. ... dave souza, talk 15:03, 29 September 2007 (UTC)
OK, so if an editor is inserting their opinion, but citing a close source that agrees with them, is that OR or not? I think that if it's previously published, it's not OR. It's using an inappropriate source per WP:V and possibly also WP:NPOV if the editor is also selectively excluding third party opinions. I think I'm starting to see what all the fuss over PSTS is about, but my conclusion is that NOR has become a WP:COATRACK for important points, that really are not fundamentally about NOR. Dhaluza 16:07, 29 September 2007 (UTC)
I get what Dave is concerned about... taking things out of context (ie not "sticking to the source") or making a synthesis based upon the source is indeed OR. I would even go so far as to say that most occurances of these two forms of OR use what the policy currently defines as primary sources. The problem is that in trying to say "don't do this" the current language seems to indicate that there are no situations where using a primary source, or a source "close to the subject" is appropriate. By focusing on the type of source, and not on the usage of the source, the policy policy becomes skewed and misinterpreted. That is the what I am concerned about. Perhaps part of the problem is that we focus so much on the ways such sources are improperly used... and don't discuss the ways they can be used properly. Blueboar 16:24, 29 September 2007 (UTC)
Not sticking to the source is OR. But what about sticking to a close source without synthesis? If someone puts self-congratulatory material from an autobiography in a Wikipedia biographical article, that's a problem, but not an OR problem, is it? It is advancing a position, but not an unpublished one; nor is it a novel interpretation of history, since it is probably rather predictable. It would seem to be out of the scope of the NOR definition. Dhaluza 18:20, 29 September 2007 (UTC)

What about when the source is the interpretation? For example, foundational texts of political movements, and later texts which are not commentaries on earlier texts. These are almost all primary sources, and are mostly composed of interpretations. Jacob Haller 18:57, 29 September 2007 (UTC)

These can only be used as sources about themselves, but this is not an OR issue, it's an NPOV issue. We can say that the organization's self-stated purpose is... then quote from these texts, and it would be completely appropriate. The issue there is putting the close source in context. But taking a published source out of context is not OR, publishing new material is. Dhaluza 22:51, 29 September 2007 (UTC)
I disagree with your last comment... taking a published source (or as is more common, part of a published source) out of context is definitely a form of OR. Doing so gives the text being used a new meaning and interpretation. However, this form of OR can be done with any "type" of source and is not a primary or secondary issue. Blueboar 13:10, 30 September 2007 (UTC)
Yes, taking a source so far out of context to the extent of using it to say something it does not is OR. That is going beyond the source. But simply using the source in an incorrect context is not OR. It's a passive mistake, rather than an active one. For example leaving the qualifier "self-stated" out of "the organization's self-stated purpose is..." in the example would not be an OR problem, it would be an NPOV problem. The description was previously published, not made up by the editor. But if the editor took part of that, and part of something else, and then put them together to say something neither said on its own, that would be an OR synthesis. The main point is that using a source to say something it actually says, but without a neutral context is an NPOV problem; using sources to say something they do not say is OR. Dhaluza 19:38, 30 September 2007 (UTC)

I just want to speak in favor of Dhaluza's observation above, that a main problem with the primary-secondary-tertiary model is its lack of granularity. The primary-secondary theory is not the right tool to express what we mean about OR, because it isn't specific enough. Classifying a source as either primary or secondary doesn't help us, because:

  • Sources can be both primary and secondary, depending on how they are used. The PSTS model defines sources in a monolithic way that is not sophisticated enough to deal with use-based distinctions. Example: Which source is more "primary"?:
  1. a diary entry analyzing raw data from a published reference work.
  2. statistical data derived from an analysis of thousands of peer-reviewed journal articles.
  • The "primariness" of a source is not what leads to OR. OR occurs when the editor goes beyond the source to inject new research, which can occur when a source is used either in a "primary" or a "secondary" way. Example: Which research is more original, and does primariness have anything to do with it?:
  1. a crackpot editor's alternative physics theory based on his own unpublished UFO sightings?
  2. a crackpot editor's alternative physics theory based on the UFO sightings (a secondary source) published by David Hasselhoff?
  • "Primariness" has implications that go beyond OR. The danger of NPOV violations is arguably lower for primary sources than for secondary sources. Example: Which is better:
  1. a peer reviewed article announcing primary conclusions on the second law of thermodynamics, or
  2. some evangelical pastor's interpretation of a peer-reviewed article on the second law of thermodynamics?

COGDEN 21:36, 1 October 2007 (UTC)

If PSTS was moved, is this policy weakened?

Once again it appears that we are going around in circles and getting sidetracked on side-issues. So, in an attempt to gain further clarification on just this one point, I would like to ask the following questions so that we can address this one point. wbfergus Talk 12:30, 29 September 2007 (UTC)

  • 1.) If PSTS was moved from this policy (not removed from Wikipedia) to some other (as yet undetermined) 'home', how would this policy be weakened? Some examples would be great.
  • 2.) Does moving a very brief definition of primary, secondary and tertiary sources out of this policy suddenly allow original research? These definitions would still be available elsewhere on Wikipedia.

If PSTS was moved comments

I still am unconvinced that we need to classify sources as primary or secondary to evaluate OR. We can simply evaluate them to a more descriptive standard, and avoid any argument over an arbitrary primary/secondary dichotomy--just say what it is, and nothing more. My first impression of PSTS was "why is that here?" and I still have that question unanswered. I don't think moving the definition out of NOR will weaken it, I think it will strengthen it by avoiding unnecessary diversions around source typing, and allow a more precise focus on OR. Dhaluza —Preceding signed but undated comment was added at 13:00, 29 September 2007 (UTC)

Yes, it would be weakened. Someone else has summarized this policy as "do not use sources inappropriately." I have made it clear consistently that at stake here is indeed how sources are used. But the point is this: it is not enough to say use sources appropriately. Different kinds of sources should be used in different ways or put proscriptively different kinds of sources should not be used in certain ways. To explain this clearly we must differentiate between types of sources. NutThus it is always two issues: types of sources, and how to use them appropriately? Slrubenstein | Talk 14:31, 29 September 2007 (UTC)

Can you elaborate on these points please? Can't rebuttals to 'OR edits' be plainly stated as "This is a statement of fact, from a Primary source, and these interpretations (or conclusions, etc.) are not made in the statement of fact. In order to include these in an article, they must have been previously published by a reliable and verifiable third-party."? Somehow (I am rather dense at times) I still fail to see how the definition of the types of source directly relate to whether something is original research or not. If I cite an article that explicly states "...yada, yada, yada, red hair occurs naturally in caucasions...", in a 'secondary source', I still cannot make an edit that states "Most caucasions have red hair". That would clearly be OR, but the type of source had nothing to do with it. I guess that is where I (and others) are getting lost in the discussions. In this case, 'source-typing' has nothing to do with whether the statement was OR, as by several of the previous discussions over the last month or so would seem to indicate the 'usual' criteria for determining OR is if the material came from a primary source, and in this (bad) example, the material was clearly a misapplication of information from a secondary source. PSTS had no bearing, and would only confuse the issue, not make it more clear. wbfergus Talk 14:50, 29 September 2007 (UTC)
A good question which requires widespread and serious consideration before any moving takes place. I've just pointed out above a way in which policy would be weakened, and the response was that this hadn't been thought about, but we could rewrite part of WP:V to compensate. What other effects haven't been thought about? Of course all this concern about which policy this part should fit in would be overcome if we amalgamate the policies into WP:A. Seems a good answer to me. ... dave souza, talk 15:17, 29 September 2007 (UTC)
I agree we need to carefully consider the alternatives, and the best one may not have even been discussed yet. Each option has costs and benefits, including the do nothing option. I think it is unfortunate the WP:A effort failed, but it is too soon to revisit it now. We are left to address the conflicting definitions and provisions of the content policies within the current framework, and I think serious discussions about what belongs where are important. At least you are willing to engage in thoughtful discourse on the individual points. The dismissive comments and blanket denials thrown around are not helpful to understanding what is important and why. The idea that everything is just fine the way it is ignores the fact that all of these policies have developed over time, and there is no reason to believe that they just happen to have achieved a perfect state at the current time. Dhaluza 16:16, 29 September 2007 (UTC)
  • Not only would the policy be weakened by the removal of PSTS, in many applications it would be flipped over backwards. Remember, the original purpose of WP:NOR was to prevent people with crackpot interpretations of sources such as mathematical theorems, physics formulas, historical documents and such. Within months it was expanded to cover all topic areas, because its implications pretty much covered the entire wiki. A year ago it was expanded from "primary and secondary sources" to PSTS to more effectively conceptualize where encyclopedias and such fit into the picture. PSTS instructs WP users who use primary sources to write about the topic in a way that "anyone—without specialist knowledge—who reads the primary source should be able to verify that the Wikipedia passage agrees with the primary source. Any interpretation of primary source material requires a secondary source." Thus, PSTS explicitly counters the almost inevitable argument that in such cases it's best to go back to the original sources as a basis for writing about obscure or debatable topics or issues. ... Kenosis 17:29, 29 September 2007 (UTC)

Kenosis, sorry if I appear to be nit-picking issues with you, it's not my intent, it's to clarify the position you and others seem to have, but you seem to elaborate a bit more, so it's more 'productive' in my mind to ask you than perhaps some of the other responders. First, this is not about removing PSTS, but just moving it, if and when a more appropriate place is found. By moving it somewhere else (probably into either another policy or a new policy, to alleviate the enforcement concerns you and others have expressed), PSTS is still an inherent part of Wikipedia and policy. Instead of defining a primary source (extremely briefly and incompletely) and the other sources, wouldn't this be better handled with PSTS somewhere else, where it could be expanded to cover the various usages in the different disciplines, etc., and this policy could just say "For the current definitions of the types of sources, see XYZ". Something along the lines of using the "main article" template. From your edits in the Sandbox, I know you've seen how contorted the various definitions are according to discipline; and as written, PSTS doesn't cover everything. If it doesn't cover everything, it can't be accurate or all inclusive, so wouldn't moving it somewhere where it could be expanded to cover everything (as the current editors and their sources currently define it) be more logical? So, with PSTS only moved and expanded to become more accurate, not deleted, how does this weaken this policy? wbfergus Talk 10:34, 30 September 2007 (UTC)
Kenosis, would you agree with the following proposition:
"anyone—without specialist knowledge—who reads the secondary source should be able to verify that the Wikipedia passage agrees with the secondary source. Any interpretation of secondary source material requires another secondary source."?
If you agree with the above sentence, then maybe you can see a point that has really been missed in this argument thus far: what you are proposing is actually a weakening of the OR policy. As applied generally in Wikipedia, NOR applies both to primary and secondary sources. The language you cite above actually weakens the NOR policy by omitting its express application to secondary sources. COGDEN 22:04, 1 October 2007 (UTC)
    • I went back and checked the article history. I think it's also important to note that until March 2005 it said Wikipedia is a secondary source, then it was changed to say that Original research is research that produces primary sources or secondary sources. The definition used to support this was: "Primary sources present information or data, such as archeological artifacts; photographs; historical documents such as a diary, census, transcript of a public hearing, trial, or interview; tabulated results of surveys or questionnaires, records of laboratory assays or observations; records of field observations. Secondary sources present a generalization, analysis, synthesis, interpretation, or evaluation of information or data." So the original definition of PSS in NOR was based strictly on factual vs. interpretive sources, and the rest of the coats were hung on this rack over time. Dhaluza 19:20, 29 September 2007 (UTC)
  • Yes, it would be weakened, and weakened considerably. ≈ jossi ≈ (talk) 17:33, 29 September 2007 (UTC)
  • I don't see why it would weaken it, perhaps someone could explain. NOR is a little confusing to new readers and not the first place one would think to look when trying to figure out if a source is appropriate. Perhaps by posing PSTS as a choice of source issue rather than an original research issue, it makes more sense so people wil follow it more readily.Wikidemo 17:58, 29 September 2007 (UTC)
  1. Yes the policy would be weakened: a) because PSTS works in tandem with WP:SYN. b) because PSTS defines how a primary source (eg an episode of a TV programme in an article about that episode or the series) can be used without becoming original research. Removing PSTS from here is not helpful.
  2. By not explaining how to use a primary source appropriately there will be confusion about what is and is not defined as OR. IMHO the definition of a policy should be as detailed and centralized as possible. Again moving or removing PSTS is not helpful.
  3. I would rather suggest that a complimentary section should be created at WP:RS to explaining how reliable certain sources are. WP:NOR needs this here, without it the policy is not 100% clear and is not comprehensive. Having information here on sources is obvious to me. By definition if an editor is using primary sources (e.g evidence or court rulings in an article about a crime that were not commented upon by journals/books/articles/documentaries about the crime) then they are conducting original research.
    The fact that new users find NOR confusing is an issue, but to my mind a separate one. If somebody could write a concise and to the point lead and nut-shell that would help newbies considerably. However, removing PSTS from NOR is unhelpful for the project because it introduces a new loophole in policy. IMHO this policy can't exist without defining what is a primary source and how to use one--Cailil talk 18:32, 29 September 2007 (UTC)
I can't understand this claim that "if an editor is using primary sources ... then they are conducting original research," but it has already been used to prevent the use of perfectly good primary sources. Of course, you are using a much narrower definition of PS than PSTS does, and PSTS uses a much narrower definition of PS than many editors do. Jacob Haller 18:52, 29 September 2007 (UTC)
Jacob, if an editor is using primary sources in ways other than to describe obvious facts then the editor is involved in novel thought. For example if an editor goes to a play and then says "the performance was terrible" using the performance as a primary source - that's novel thought. But if they find a review of that performance (a secondary source) saying the same thing and says The New York Times review described the performance as "terrible" - that's not novel thought. Now I don't see how I'm using a stricter definition. Primary sources are sources "very close to the" the subject of the article. This is the standard definition in academia. This has been the situation on WP for as long as I've been here (reliable sources are secondary sources independent of subject - WP:RS WP:N).
I do think the wording of PSTS should be looked at. But the spirit of rule, the idea that WP is not here for novel interpretations, that WP is here to present verifiable, mainstream knowledge should not change - we wouldn't be wikipedia any more if it did--Cailil talk 19:33, 29 September 2007 (UTC)
Actually, primary sources were originally defined as factual, rather than close sources. That bit was added later. Dhaluza 19:42, 29 September 2007 (UTC)
It's always been factual in some important sense of the word "factual", and remains so today. Where the facts revolve around what was actually written in a leading philosophical treatise or a literary work such as a novel or play, that's primary, and broadly speaking it's factual about what was written. When it involves interpretation beyond the plain meaning of the words, it's interpretive, or secondary. ... Kenosis 20:46, 29 September 2007 (UTC)
Where in WP:RS does it specify that only secondary sources are reliable? SamBC(talk) 20:25, 29 September 2007 (UTC)
Sorry Sambc - I quoted the wrong policy - I was referring (& should have said) to WP:N--Cailil talk 20:49, 29 September 2007 (UTC)
Ah, and there it becomes a little irrelevant. WP:N has the additional requirement of being secondary purely for the purposes of notability. Sources establishing notability should generally be secondary, but no policy says that a source must be secondary to be a reliable source of factual information, that I'm aware of. Plus, WP:N is a guideline, not a policy, as in fact is WP:RS. SamBC(talk) 20:55, 29 September 2007 (UTC)
I take your point that WP:N is a guideline but while the quote is from WP:N the point is taken from WP:RS - independent or "third party" published sources. Primary sources cannot be independent of the subject (ie third party). Yes primary sources can be used to describe unambiguous/obvious facts (ie the names of characters from a TV show are obvious from an episode of the TV show and so an episode can be sourced) and yes they have a place on WP but an article or section composed only (or mainly) of points made from primary sources is original research. That sort of writing is an essay, a piece of original argument--Cailil talk 21:46, 29 September 2007 (UTC)
  • If it is removed now - perhaps weakened in some aspects, perhaps strengthened in others. The current phrasing encourages edit-warring, discourages appropriate use of primary sources, and encourages synthesis from secondary sources, for editors who think in terms of primary vs. secondary instead of sticking vs. synthesis.
  • If it is removed after Blueboar's addition, or another similar addition - perhaps weakened in some aspects, certainly strengthened in others.
  • With suitably-written additions we can make PSTS redundant, and move PSTS elsewhere without any trouble. Jacob Haller 18:52, 29 September 2007 (UTC)
Could you provide diffs of the edit warring? It's not something I can recall. .. dave souza, talk 19:46, 29 September 2007 (UTC)
  • P.S. Why does "any interpretation of primary source material" require a "secondary source?" Many primary sources make their own, perfectly clear, completely unambiguous, interpretive claims. Nonetheless, many editors interpret PSTS to justify the removal of interpretive claims found in the primary sources. Jacob Haller 19:03, 29 September 2007 (UTC)
As stated above, attribute the interpretation or claim to the source and make it a fact that they hold that opinion. .. dave souza, talk 19:46, 29 September 2007 (UTC)
I don't see much difference between "John Doe argued X [ref John Doe]" and "According to John Doe, John Doe argued X [ref John Doe]."
As an example, see Talk:Benjamin Tucker#Spurious_Edits. [29]. Note that beyond POV-pushing, those edits add others' descriptions of Tucker's views and remove Tuckers' own descriptions of his views.
Intangible argues, in defense of the removal of Tucker's self-description: "Also read Wikipedia:No_original_research#Primary.2C_secondary.2C_and_tertiary_sources. One should only use quotes of Tucker himself in a descriptive way, to illustrate a point made from a secondary source. Intangible2.0 11:51, 25 April 2007 (UTC)"
Full Shunyata argues, in defense of the inclusion of Tucker's self-description, as well as the removal of Yarros' description: "The quotes from Tucker are used in a descriptive way. They are describing his views on Anarchism and his views on the various topics discussed in the paragraphs I put them in. For instance, I quoted his take on labor unions because labor unions were the topic at hand in the paragraph I inserted the quotes. The first couple of sentences in that paragraph claimed that Tucker was opposed to labor union actions. Which is not entirely true as can be seen from his own words in support of labor unions. And very few of my quotes are from secondary sources, many are directly from Tucker. What makes Victor Yarros and Wendy McElroy so authorotative on the subject? Are they God? Are they more authoritative on the subject of Benjamin Tucker than Benjamin Tucker himself? Full Shunyata 15:48, 25 April 2007 (UTC)"
Remember, we're talking about Tucker's beliefs, not his actions; people can boast about their actions, so we may want secondary sources there, but it's hard for people to boast about their beliefs.
This was a complicated dispute involving neutrality and undue weight issues as well as different interpretations of NOR. Nonetheless, I think the edit war on the article and the dspute on the talk page attest the misuse of PSTS to remove legitimate primary sources. Jacob Haller 21:15, 29 September 2007 (UTC)
And the point is what? That those writing the article about American anarchist Benjamin Tucker are unable work out the content disputes because of the way WP:NOR is presently written? The language of the section on PSTS is adequately clear that these users have an understanding of what the primary sources are in relation to Benjamin Tucker's own writing. The editors at that article are permitted to use Benjamin Tucker's own writing to convey what Tucker wrote (within practical limits of a brief article, of course), but only if any reasonable person could tell what he meant on a plain reading. If interpretation is needed, use the secondary sources. If there's disagreement about interpretation, use the secondary sources, in keeping with WP:NPOV#Undue_weight. If something said or written by Benjamin Tucker is argued by other editors to be taken out of context, use the secondary sources. ... Kenosis 21:43, 29 September 2007 (UTC)
I think that is exactly the point. A conflict about the proper use of the subject's quotes in an article about the subject, became an argument over primary and secondary sources in relation to NOR. This was an unnecessary distraction, and only compounded the original problem. Dhaluza 23:16, 29 September 2007 (UTC)
Up at least until September 11, 2006 this policy still had the following 4 sentences, though it appears it was beginning to change around this time, I'm still browsing the edit history to see why it finally disappeared:

Original research that creates primary sources is not allowed. However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged. All articles on Wikipedia should be based on information collected from published primary and secondary sources. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia.

[30]
It appears that many of the last months edits and discussions wouldn't have happened if that paragraph was still here. wbfergus Talk 22:04, 29 September 2007 (UTC)
I would support a reinsertion of something like this paragraph, or even a verbatim reinsertion of the paragraph, so long as the PSTS analysis is kept intact. Many content disputes have been resolved throughout the wiki based on the primary/secondary distinction, and the addition of tertiary helps in many situations where the secondary sources need integration. The rest is up to the consensus process at the individual articles. ... Kenosis 00:37, 30 September 2007 (UTC)
It appears that this paragraph diasappeared on this edit diff]. I can't imagine that it was done intentionally, but as much as this 'section' was being discussed during this timeframe, I find it equally unimaginable that nobody noticed it's disappearance. I can't find anything on it in the relevant talk pages either. It looks like there were so many edits and discussions, that this somehow slipped through the cracks. wbfergus Talk 22:45, 29 September 2007 (UTC)
Yes, it also appears that that diff is the one that first introduced the conflation of close and factual sources as well. I think putting back the lost clarification on primary sources, and removing this close source description, would be a very good start at cleaning-up NOR. Dhaluza 23:04, 29 September 2007 (UTC)
It's not a conflation. Whether it's "close to the situation you're writing about" or "factual" depends on the situation being written about. No doubt a lot of people were confusing "factual" in a way that excluded facts about fictional writing, facts about original philosophical works, etc. It depends on the particular topic being written about. A situation like this was just mentioned above. Benjamin Tucker's writing was largely theoretical -- in this case "factual" means "just the facts about what Tucker wrote or said", and "close to the situation you're writing about" means "Benjamin Tucker's published writing, speech, personal notes, etc." ... Kenosis 02:35, 30 September 2007 (UTC)
I'm not sure what you are objecting to. If you don't like the word "conflation" then replace it with "juxtaposition". But if you are trying to justify the conflation, then I completely disagree. Whether a source is close or not is a different concept than whether it is used to support a statement of fact or opinion. The two are mixed in the current definition which was substantially changed in [this edit]. That may have improved the definition of primary source, but it obscured the meaning of original research. Whether a source is close or not is not central to NOR. If it was published by a reliable source, it's not OR. If the source is not reliable in the context it is used in, then it's a sourcing problem. NOR is a non-sourcing problem. Dhaluza 09:26, 30 September 2007 (UTC)

I also agree that we should put those sentences back into the policy. My concern here has not been mainly that "primary" is unneeded, although I believe that to be true; my main concern is the idea that has taken hold among some that what experts say is of necessity a primary source (SlimVirgin said "looking around to see how mainstream sources use terms is exactly what we're supposed to do when there's a dispute about how to use them" when I said "Just because an expert says it, doesn't make it a primary source" in a discussion held at Wikipedia:Articles for deletion/Challenges and issues of industrial agriculture) and that secondary sources trump primary sources (indicated by SlimVirgin in the battle for the definition of "factory farming" at Factory farming; see the first sentence in that article and its sources and their use for what SlimVirgin and friends think is the proper use of sources - which I believe to be OR - but which they insist is supported by this policy's statements about primary and secondary sources); both positions being enforced by SlimVirgin's editing of this policy. My main concern is the idea that scientifically preferable sources are thus downgraded compared to newspaper accounts which would hurt the evolution article and not help it and leads to absurd conclusions like the saying Wikipedia Media Group is Wikipedia's parent company [31] according to Diligent Media Corporation ("[a] is a joint venture [... w]ith a reach of more than 120 countries and access to more than 250 million viewers globally [...] with presence in major global markets, including Asia-Pacific, the Middle East, the UK, the US, Canada, the Caribbean and Africa"[32]) which is thus a reliable published source providing a notable third party evaluation of Wikipedia which can not be dismissed due to primary sources (but could be dismissed with enough other secondary sources that say otherwise rendering it too insignificant a POV to necessitate its inclusion). WAS 4.250 08:14, 30 September 2007 (UTC)

I also agree with putting those sentences back into the policy, and that they would go a long way towards solving the problem. They make it clear that primary sources can be used. --Coppertwig 00:48, 1 October 2007 (UTC)

I think they should go back in too. I think other discussion is leading me to a view that if we could detatch PSTS from being about NOR and restate it as a means to evaluating all policies then I think we might work something up which is far more convincing. Spenny 22:05, 1 October 2007 (UTC)


  • Moving, and in fact deleting, the PSTS model would actually strengthen NOR. Why? because the PSTS model says,
anyone—without specialist knowledge—who reads the primary source should be able to verify that the Wikipedia passage agrees with the primary source.
This is way too weak! We're giving secondary sources here a free ride. Here's how it should have read, if we really wanted to prevent OR:
anyone—without specialist knowledge—who reads the primary or secondary source should be able to verify that the Wikipedia passage agrees with the primary or secondary source.
Absent the present weak language, the NOR policy prevents any type of going beyond the sources, whether the sources are primary, secondary, red, green, Spanish, extraterrestrial, pornographic, or otherwise. Why do we want to weaken that policy by making distinctions based on the type of source? COGDEN 22:20, 1 October 2007 (UTC)

Yet another proposal (yes, again)

Okay, in light of the general mix of comments in a preceeding section, would there be agreement to add to the policy the following paragraph (I took the liberty to add one more sentence)?

Per WP:NOT#OR, Wikipedia is not a primary source. Original research that creates primary sources is not allowed. However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged. All articles on Wikipedia should be based on information collected from published primary and secondary sources reliable published sources. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia.

And if there is agreement (consensus) to add the above paragraph, where should it go? I think right at the beginning of the "Sources" section, before "Reliable sources", would be the logical place. wbfergus Talk 10:09, 30 September 2007 (UTC)

Agree -

Oppose -

Comments -

  • I would rather see us define OR without using the primary/secondary source distinctions if they are not necessary. So I would rather see it introduced as "Wikipedia does not bear witness to any fact" Dhaluza 11:29, 30 September 2007 (UTC)
  • I have a feeling that the sentence: "However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged" may need to be expanded. This is a great discription of souce based research, and if done properly this should indeed be "strongly encouraged"... however, if done improperly "collecting and organizing information" can easily result in synthesis. Blueboar 13:31, 30 September 2007 (UTC)
    • I don't think that this is strictly true. Collecting and organizing information is fundamentally different from synthesis. Selectively collecting and presenting items is a NPOV problem--synthesis requires cognition. Everyone has unconscious tendencies that might prevent them from seeing certain things, or viewing the things they do see in a certain way. But if the person is not consciously pushing their POV, I would say it is not OR. This is just like a computer performing tasks according to an existing program--it may lead to a biased presentation, but that is a different problem best solved by involving more editors with different programming. OR synthesis requires original thought, and is a more active process. It requires someone who thinks their POV needs wider dissemination, and sets out to use WP as a soapbox to project their views. Rather than simply favoring one set of pre-published views over another, this POV type OR requires assembling evidence and publishing new views based on a novel interpretation or re-interpretation of it. So someone bent on synthesis must collect evidence to synthesize, but simply collecting evidence does not necessarily lead to synthesis. Dhaluza 13:55, 30 September 2007 (UTC)
  • I agree with Blueboar - the piece about organizing sources needs more detail. If we retain the existing points in policy that "Primary sources that have been published by a reliable source may be used in Wikipedia, but only with care [...] Any interpretation of primary source material requires a secondary source." - then this addition is fine. However, I feel that the use of the phrase "Original research that creates primary sources is not allowed" will be confusing. I would prefer to say - "Original research consisting of novel synthesis or interpretation of sources is not allowed"--Cailil talk 13:52, 30 September 2007 (UTC)
  • (ec) I strongly recommend that "primary and/or secondary sources" be replaced with "reliable sources". Otherwise, I will have to strongly oppose, due to numerous problems with the phrasing making it easily abused, such as the implication that sole reliance on primary sources is acceptable. Vassyana 13:54, 30 September 2007 (UTC)
  • Are there any comments or concerns with changing the beginning from "Wikipedia is not a primary source. Original research that creates primary sources is not allowed." to "Wikipedia does not bear witness to any fact. Original research must first be published in another reliable source, which is then cited to support its inclusion in a Wikipedia article." This avoids the need to pre-define primary source, and allows this statement to precede the PSTS section. I think it is also a stronger statement because it focuses on the use rather than type of source, and says what to do, rather than what not to do. By saying what it means directly, it can easily be interpreted by someone without relying on specialist knowledge of source typing. Dhaluza 14:14, 30 September 2007 (UTC)
    • I would suggest (though it looks ugly) using strikeout for any changes. This way, once others join in this discussion (see below), they can more readily see any changes from the original. wbfergus Talk 14:30, 30 September 2007 (UTC)
  • I have no objections to the strikeout and new wording. However, we still need to hear from the other respected members who may still have some valid objections, notably Jossi, Slrubenstein, and Kenosis. There are also others who have been active in these 'discussions' who haven't weighed in yet, so let's not get to hasty. wbfergus Talk 14:13, 30 September 2007 (UTC)
  • The earlier language was significantly clearer, IMO. it read

Original research that creates primary sources is not allowed. However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged. All articles on Wikipedia should be based on information collected from published primary and secondary sources. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia.

There's no need for the words "Per WP:NOT" because the particular policy emanates from this page, not WP:NOT. And the language "Wikipedia is not a primary source" is potentially misleading at this stage of development of the project. Actually, WP should never be a primary source, and when it is a secondary or tertiary source, it should be devoid of original analysis or interpretation. ... Kenosis 16:31, 30 September 2007 (UTC)