Wikipedia talk:No original research/Archive 31

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 25 ← Archive 29 Archive 30 Archive 31 Archive 32 Archive 33 → Archive 35

Revisiting use of primary sources in controversial articles

In a discussion above, the comment was made by COGDEN: Also, since I work a lot in highly-controversial history fields, I know from experience that the greater the subject matter's controversy, the greater must be the reliance on primary sources. You don't really understand this until you work on articles in ultra-controversial fields. I'd like to revisit this discussion because that's perhaps a major lynchpin to the deadlock here. If that's the real thrust behind trying to rewrite this policy, then it should be dealt with head on first, without any further playing around with the words used. In the context of my understanding of the existing policy, from my experience in disputes in controversial articles, and from my admittedly somewhat pedestrian background in formal academic writing, this idea of bypassing claims made in the secondary sources in favor of claims from primary sources would an example of using primary sources for novel claims about the topic, ie "original research". If they weren't "novel claims", then there would be no need to deliberately bypass the secondary sources anyway. Am I missing something? Professor marginalia (talk) 23:43, 12 December 2007 (UTC)

Yes and no. :)
Cogden's supposition that "the greater the subject matter's controversy, the greater must be the reliance on primary source," though appropriate in a secondary source, is not desirable in an encyclopedia. We don't do research ourselves.
This is not however the "lynchpin to the deadlock," nor do I think it is "the real thrust behind trying to rewrite this policy."
The lynchpin to the deadlock is that the PSTS section is ridiculously wordy, to extend to the complete superfluity for any differentiation of "primary" versus "secondary". When sources are just sources, and the properties of "original research" are properly defined, there is no need for the junk.
The pitiful prose and internal contradictions are secondary; they are an effect of the desire/need to split hairs (which get split finer and finer and finer...).
-- Fullstop (talk) 00:39, 13 December 2007 (UTC)
Nobody should ever be able to introduce novel claims and theories into an article. Bypassing the secondary sources means, in most cases, simply quoting or otherwise-non-controversially-citing the controversial primary source, or only making the points made by the original author, without further commentary or spin. Quotation is often the best, but even direct quotations of the primary source runs afoul of the present PSTS language, if what is quoted is an analysis or conclusion by the original author. COGDEN 08:50, 13 December 2007 (UTC)
Cogden's position on this issue is simply common sense. The Easter article used to be plagued by edit warring over the origin of the name. People would write that it was derived from the name of an Anglo-Saxon goddess, Eostre, and cite various reliable secondary sources stating this. Then someone else would come along and change the article to say this was a myth and no such goddess was ever worshiped by the Anglo-Saxons... and they would cite just as many reliable secondary sources establishing that claim. The dispute was only settled when people went back and cited the primary sources and documented the reasons for the disagreement... namely that the earliest source, Bede, had very little information to go on and may simply have gotten it wrong... or not.
Information which is or may be disputed should always be referenced to a primary source, so that people can go and see for themselves what the actual evidence is... rather than relying on someone else's secondary source interpretation. We rely on secondary sources for interpretative commentary, but when interpretations differ the primary sources are essential to understanding of the dispute. --CBD 14:26, 13 December 2007 (UTC)
This is a misinterpretation, or misrepresentation, of PSTS. The use of primary sources, for example in the "Easter" article just mentioned, should be limited to the scope permitted under PSTS, which is that Wikipedia users should not make any analytical claims about Bede's writing or whatever the "primary source(s) are considered to be in that topic, and should use that source only in a way that any reasonably educated person can double-check the souce and conclude that WP's reporting of that source is accurate without possessing specialist knowledge. Do people bother to read the policy? ... Kenosis (talk) 14:34, 13 December 2007 (UTC)
It could have and should have been relatively easily solved by simply documenting the various and sundry notable views of the matter. That's exactly what is demanded by the fundamental policy of neutral point of view. Also, citing the primary source in the fashion you mention sounds exactly like original research. We should be depending on secondary sources to tell us that Bede is the origin of the dispute/conflict and that he had little information to go on and therefore made incorrect assumptions. Making such an analysis ourselves is certainly original research. Vassyana (talk) 14:41, 13 December 2007 (UTC)
I might also mention that the article does not seem to rely on Bede directly, but rather reports what secondary sources stated about the matter ... exactly as it should be. Vassyana (talk) 14:45, 13 December 2007 (UTC)
Exactly, Vassyana. The Easter article is an example of why editors need be careful about their use of such sources. Any policy wording which encourages we nobodies (ie WP editors) to cite an 8th century manuscript, written in Latin no less, to settle out or override conflicting claims in good secondary WP:RS is insanity. And a check of the Easter article shows how good editing works here--secondary sources are used to give a fairly well developed discussion of the differences of opinion among scholars about Bede's account. Even though it's hard to keep track of what the PSTS actually says anymore, I can't recall any version of it so far proposed which would disallow editors from quoting Bede. Use is allowed "with extreme caution" -- PSTS policy is excellent advice for the etymology of Easter. Professor marginalia (talk) 15:24, 13 December 2007 (UTC)
No, look carefully. The article does quote Bede directly, then it provides secondary commentary to Bede. This is a good example of how primary sources should be used. COGDEN 02:39, 14 December 2007 (UTC)
Kenosis, how is what I described NOT 'limiting the use of primary sources to the scope permitted under PSTS'? The use of primary sources on the origins of the name Easter is entirely consistent with this policy. As to Vassyana's claim that the matter could have been handled, "by simply documenting the various and sundry notable views of the matter"... that's precisely what WAS done, but you can't DO that without citing the primary sources that people are disagreeing about. Otherwise you are simply saying, 'some people think this, but other people think that' without providing any context. You lot seem to be disagreeing with me as vocally as possible... while advocating exactly the same thing I am. --CBD 16:33, 13 December 2007 (UTC)
The article doesn't cite the primary source. It cites secondary sources that discuss the primary source. I disagreed because your description sounded like original research, as I mentioned above. Vassyana (talk) 16:42, 13 December 2007 (UTC)
Ok, then apparently we disagree on the meaning of the word 'cite'. I see the article quoting De temporum ratione, and call that 'citing the primary source'. You apparently don't... it's 'quoted', but not 'cited'? --CBD 17:05, 13 December 2007 (UTC)
The quote is given context entirely by the secondary sources and the controversy is entirely cited to secondary sources. The quote is nice to point out what is being discussed exactly, but it's hardly essential to the article. The secondary sources do all the heavy lifting and the primary sources serves as nothing more than illustration. That's certainly not the picture you painted. Vassyana (talk) 17:11, 13 December 2007 (UTC)
To prevent misinterpretation, let me be clear that the primary source quotation serves its purpose well as providing illustration to the discussion and that its use in the article is not at all inappropriate. On the contrary, the article is a good example of how a primary source should be used (complimenting its secondary source discussion). It's just that it is not at all used in the way you described, nor does it have the central place in the article material that you expressed. Vassyana (talk) 17:14, 13 December 2007 (UTC)
Bede is cited, as he should be. Controversial interpretations of Bede are also cited directly, as they should be. We see Bede first through Bede's own words, then there is the discussion of various commentators. This violates PSTS in another way, incidently, because a lay person who does not know Latin cannot verify it. Is that bad? No. It happens all the time, and is allowed under WP:V. COGDEN 02:39, 14 December 2007 (UTC)
Eh, I see the Bede quote as nothing more than illustrative (though it serves the purpose well). However, knowledge of Latin is most certainly not needed to verify the passage. Copies of Bede in both Latin and English translation are easily available, even online. Regardless, the use of foreign language sources and user translations is a long standing permissable practice in Wikipedia and is not indicative of a broader rule. Though I admit I strongly disagree with this exception to the general rules of verifiability and no original research, it is a long-standing exception even if occasionally controversial. Vassyana (talk) 05:29, 14 December 2007 (UTC)

(outdent) Reply to Cogden's of 02:39, 14 December 2007. For a while now, I've been worrying about your and other's reading of these two passages in PSTS:

and

in relation to hard-to-comprehend subjects such as philosophy, nuclear physics, and now Latin.

You appear to be interpreting the words to mean that the user must be able to understand what the primary source passage actually means. But, in my view, that's neither the intention nor is it what the words actually say. What they say is that the average user should be able to verify that the WP passage agrees with the primary source. Well, even without knowing Latin I could easily verify that "Eostur-monath, qui nunc paschalis mensis interpretatur..." agrees with "Eostur-monath, qui nunc paschalis mensis interpretatur..." in the published primary source.

Maybe the Bede quote is not the best example of what I'm talking about because it is somewhat muddied by the issue of translation - in my view the source of the translation should be separately cited because a translation from Latin isn't something that the average user could be expected to verify, so it isn't an appropriate inference to make from the primary source.

But anyway direct quoting like that is at one end of the spectrum. Restriction to direct quoting of statements made by Derrida would probably be a good idea too, from what I've read. At the other end of the complexity spectrum lies simple text which would be far more amenable to rephrasing. Take this diary entry for instance:


It would be easy for anyone to verify that a WP passage that stated: "After passing the last town they lost a large piece of tread from a tire and later the gear shift lever broke off." agrees with this source.

Most primary sources will fall between the Bede and the diary extremes and PSTS tells us to use an appropriate amount of care when using them. As the meaning in the source becomes more complex or otherwise unavailable to the average user, it becomes more important that the WP passage follows it closely, culminating in direct quotation if necessary.

Does this make sense? I wonder if the problem is the use of the word verify - perhaps confirm would have less baggage?  —SMALLJIM  13:14, 14 December 2007 (UTC)

The above doesn't just apply to primary sources, of course. If it's right, it's a general principle for the citation of any source.  —SMALLJIM  13:25, 14 December 2007 (UTC)

I'd be fine with changing verify to confirm. This has been repeated over and over, but nothing in the PSTS clause disallows editors from citing primary sources-it disallows editors from making original claims regarding primary sources. The comment above said, "The dispute was only settled when people went back and cited the primary sources and documented the reasons for the disagreement... namely that the earliest source, Bede, had very little information to go on and may simply have gotten it wrong... or not." The primary source written by Bede cannot settle such a dispute. It cannot lay claim itself to being the earliest source or the best, most accurate source, nor Bede the most esteemed authority on the matter. It cannot be used by itself to demonstrate there is a dispute among historians over its accuracy. If a WP editor were to attempt any of those claims from the primary source alone, without secondary sources, that Bede is the earliest, most accurate, or questioned by later scholars, that would be original research. Take this hypothetical: What if Bede were not a widely studied subject covered in volumes upon volumes of secondary sources, and in the midst of a dispute over the origin of the word, Easter, some WP editor were to cite as a new source the 8th century Latin manuscript, the natural question should be among editors in the dispute, what secondary source do we have that can substantiate this interpretation? If there were no secondary sources of Bede, it sure looks like original research. Professor marginalia (talk) 15:37, 14 December 2007 (UTC)

If we split PSTS into definitions and usage...

Further to my second comment to Kenosis above[1] I'm being bold (for a change) and have prepared a revised version of PSTS at User:Smalljim/PSTS rewrite1. I haven't rewritten anything; all I have done is to move each of the paragraphs and bullets into one of two replacement subsections: "PS&T sources - definitions" and "PS&T sources - usage". I'm not submitting this as a finalized rewrite, or even as an alternative to Vassyana's rewrite, but it may help clarify where the disagreements here actually lie - is it the definitions, or the usage? To my mind though, it would be a more logical and tidy way of presenting the information. It also points up how little we presently say about what secondary sources actually are, which should be important because WP:N's "This page in a nutshell" directs readers here. Hope this helps.  —SMALLJIM  14:04, 13 December 2007 (UTC)

Please take a look over the explanations for the text in my draft. They may or may not be useful as reference points. Vassyana (talk) 14:32, 13 December 2007 (UTC)

::I'm confused how many drafts we're considering now. I like the ""Primary, secondary, and tertiary sources - usage" in this version, but support merging another suggestion from below to secondary sources clause so it would say something like: "Secondary sources. Where interpretive claims, analysis, or synthetic claims about primary sources are included in Wikipedia articles, use secondary sources rather than original analysis by Wikipedia editors. However, the claims can be biased by the views of their author(s) and/or publisher, so caution is required to preserve the neutral point of view. Care should also be taken to avoid undue weight and ensure the information cited is used in context." Professor marginalia (talk) 23:42, 13 December 2007 (UTC) Wrong one! Professor marginalia (talk) 14:52, 14 December 2007 (UTC)

Please don't consider this as an alternative draft for consideration to replace the existing PSTS - I'm behind Vassyana's developing rewrite for that, and I certainly don't want to start a "my version is better than yours" war! I've changed the header to remove that impression. It was simply an idea that arose out of an earlier comment about the need to maintain the primary/secondary distinction for WP:N, but (probably) not for OR. I also thought that splitting it like this may help clarify in which parts of the present section the main disagreements here lie (I think it's pretty clear that they're mostly about PS usage, with some on PS definition). I note from later comments below that some see definition and usage as being inextricably bound together, so I assume they wouldn't want to see them split in this fashion anyway.  —SMALLJIM  14:21, 14 December 2007 (UTC) Thanks- Professor marginalia (talk) 14:52, 14 December 2007 (UTC)

Grace period?

Can we delete an uncited info just immidiatly after its addition or should we have a grace period? I mean if someone added an uncited info, should't we give it the benefit of the doubt and wait some time before deleting it? If you agree on this grace period, then how long is it? Is there any policy for it or just it is up to people?Farmanesh (talk) 05:31, 14 December 2007 (UTC)

This is more of a verifiability question. There's a decent amount of tension between immediate removal and letting it ride nigh forever. Generally, unless the information included is about a living person or is blatantly wrong, it's considered a bit rude just to remove the information. If the fact or claim is completely noncontroversial (like say, the earth revolves around the sun), no citation is needed. If you think it's appropriate to challenge the claim, adding {{fact}} at the end of the statement to alert editors to the need for a reference is common practice. It would also be considerate to leave a polite message for the user who added the information requesting a source for the claim. If after a reasonable amount of time (at least a few weeks) a source cannot be found, it would be appropriate to remove the statement. If you can, find a reference for the article claim yourself. If you find a reliable source that does not completely match the article statement, go ahead and cite the source and rewrite the article claim to match the reference. Hope that helps! Vassyana (talk) 06:05, 14 December 2007 (UTC)
Thanks for the answer. But it seems very arbitary, basically it depends on the mood a person is in. As far as policy goes anyone can go around and immidiatly delete anything without citation. Policy-wise is that a fair say? If so where can we give suggestion to add this to wiki policies a reasonable grace period if it doesn't involve vandalism or live people?Farmanesh (talk) 06:40, 14 December 2007 (UTC)
See Wikipedia:Citing_sources#Unsourced_material. With respect to the grace period, Wikipedia_talk:Citing_sources might be the good place to ask. I remember reading somewhere that "a month" was sufficient, but I can't remember where I saw that. -- Fullstop (talk) 15:38, 14 December 2007 (UTC)

Secondary sources

The secondary sources paragraph of the draft seems fairly uncontroversial, and agreeable to some who object to the overall change. Basically, it would replace the current secondary sources paragraph with (ref tags no-wiki'd for ease of view):

Secondary sources are references that draw on research and other references to make interpretive, analytical, or synthesized claims, or create a general overview.<ref>[http://lib1.bmcc.cuny.edu/help/sources.html Borough of Manhattan Commmunity College, A. Philip Randolph Memorial Library, "Research Help:Primary vs. Secondary Sources"] notes that a secondary source "analyzes and interprets primary sources", is a "second-hand account of an historical event" or "interprets creative work". It also states that a secondary source "analyzes and interprets research results" or "analyzes and interprets scientific discoveries".</ref><ref>[http://www.nationalhistoryday.org/SecondarySources.htm The National History Day website] states simply that: "Secondary sources are works of synthesis and interpretation based upon primary sources and the work of other authors."</ref> Secondary sources are usually preferable because they generally provide analysis, offer a more independent view and provide a broader context for the subject. However, they can be biased by the views of their author(s) and/or publisher, so caution is required to preserve the neutral point of view. Care should also be taken to avoid undue weight and ensure the information cited is used in context.

For a view of how this looks without nowiki tags:

Secondary sources are references that draw on research and other references to make interpretive, analytical, or synthesized claims, or create a general overview.[1][2] Secondary sources are usually preferable because they generally provide analysis and a broader context for the subject, and usually offer a more independent view. However, they can be biased by the views of their author(s) and/or publisher, so caution is required to preserve the neutral point of view. Care should also be taken to avoid undue weight and ensure the information cited is used in context.

  • Trailing sentences removed, as they require a separate discussion.[2]
  • Revised based on SmallJim's suggestions.[3]
  • Added statement about using secondary sources for the interpretation of primary sources, per Professor marginalia.[4]
  • Revised based on SmallJim's suggestions.[5]
  • Draft moved to live policy. Trailing sentence removed (see subsection below). 23:01, 15 December 2007 (UTC)

The reasoning for the text may be viewed at: User:Vassyana/NOR 002/Secondary. Thoughts? Vassyana (talk) 14:51, 13 December 2007 (UTC)

Looks excellent to me. Blueboar (talk) 14:54, 13 December 2007 (UTC)
It does have the problem that the footnote is only helpful for articles on historical events or phenomena. — Carl (CBM · talk) 15:35, 13 December 2007 (UTC)
It's just the footnote from the current version of policy. I'll see if I can find another to compliment it. Vassyana (talk) 16:39, 13 December 2007 (UTC)
I've replaced the footnote with two footnotes. One that included both humanities and scientific definitions, and the other a succinct definition. Vassyana (talk) 17:04, 13 December 2007 (UTC)
Looks good. The footnote can be expanded or generalized to include other type of material that is not historical. ≈ jossi ≈ (talk) 15:45, 13 December 2007 (UTC)
Excellent Vassyana, except for the last two sentences which endeavor to replace the concept of tertiary sources. I'd like to hear opinions of, say SlimVirgin, FeloniousMonk and some other seasoned editors who've handled significant content disagreements in the article namespace that actually involved PSTS. But other than the last two sentences, I'd support it as an immediate replacement for the present paragraph on secondary sources. ... Kenosis (talk) 16:48, 13 December 2007 (UTC)
Including the last two sentences don't require the deprecation of the tertiary source category, though I admit that it is their purpose by addressing the sole substantiated point about tertiary sources. Vassyana (talk) 17:04, 13 December 2007 (UTC)
I advocate simply using it as a replacement for the current version of the secondary-source paragraph. My sense is that if it goes in without the last two sentences (after the words "neutral point of view"), SlimVirgin and FeloniousMonk would likely find it acceptable, a supposition based on prior conversations I've had with these two seasoned WP admins. If I'm wrong they can take it out of my hide, so to speak. I'll certainly defend the basic explanation that is proposed for secondary sources, maybe give or take a word or two, though that would be very minor quibbling. The overall sense of the paragraph faithfully represents the spirit of the policy, IMO. ... Kenosis (talk) 17:13, 13 December 2007 (UTC)
(ec) To be a bit cynical, it's nice that they've spoken to you about such matters, but it would be nicer if they discussed it here with the rest of the class. Without the deprecation of the tertiary category, what would be the problem with the closing two sentences? Vassyana (talk) 17:18, 13 December 2007 (UTC)
I'm only telling you my sense of it. This proposed replacement is defensible IMO to both SlimVirgin and FeloniousMonk, so long as it doesn't mess with the basic PSTS approach that is currently in place. These two very experienced admins are not out for blood, only to defend the basic policy that's been in place for quite some time because the policy has been shown to work once people understand it. That's my opinion, and if it's wrong then one or the other of these particular defenders of the longstanding policy will surely correct me. My point is to try inserting it without messing with the overall approach and see where it goes. I personally will assist in defending the proposed language, stopping at the words "neutral point of view" in the proposed draft. The reason for stopping at those words is that the last two sentences are essentially designed to take the place of what is presently written in the paragraph on "tertiary sources". SlimVirgin was the one who added tertiary sources to the "formula" over a year ago and continues to support the basic approach, and thus I believe I can defend it to her so long as the replacement paragraph doesn't negate this. FeloniousMonk, who I have had more direct conversations with in the context of articles related to intelligent design, also supports the basic PSTS approach presently in place, and the version proposed above is consistent with my understanding of why he defends the current policy. In other words, in my opinion the proposal is fine, indeed a welcome improvement, except for the last two sentences. ... Kenosis (talk) 17:31, 13 December 2007 (UTC)
No problem, I understand. Please pardon my frustration, as I should not direct it at you. I do believe they are acting completely in good faith, however it not only fails to be helpful but is downright disruptive when they will not join in the discussion and substantiate their objections. I've removed the last two sentences because they obviously require a separate discussion of their own. (They can always be tackled after we finish with this particular proposal. It's fairly clear we need to address any changes piece by piece.) I've revised the language a bit based on SmallJim's excellent suggestions. If you have further suggestions or criticisms, I'd be happy to see what we can do to address them. Vassyana (talk) 21:23, 13 December 2007 (UTC)

For what they are worth, here are my suggested alterations:

"Secondary sources are references that draw on research and other((a)) references to make interpretive, analytical, or synthetic claims, or create a general overview.[2] Secondary sources are usually the best references to cite((b)) because they ((c)) provide analysis and a broader context for the subject and usually also offer a more independent view((d)). However, they can be biased by the views of their author(s) and/or publisher, so caution is required to preserve the neutral point of view((e)). Care should also((a)) be taken to avoid undue weight and ensure the information cited is used in context. ((f))References that provide a broad treatment, such as other encyclopedias((g)) and university-level textbooks summarize((h)) large bodies of literature and research, so they can be((i)) useful for avoiding original synthesis and undue weight."

  • a. phrasing
  • b. expand for clarity
  • c. rm redundant word
  • d. reorganize sentence for clarity and style and...
  • e. sentence moved and rephrased so it flows from previous idea.
  • f. tentatively remove "review articles". Offhand I don't know what they are, so will others?
  • g. tentative addition
  • h. simplify
  • i. run-on, simplify and rephrase

Hope this helps,  —SMALLJIM  17:39, 13 December 2007 (UTC)

I must say that your suggestions and feedback have been consistently well-thought, constructive and helpful. Vassyana (talk) 21:23, 13 December 2007 (UTC)


--
Taking all so far noted comments into consideration, these are my suggestions,...
(matches original points but not original sentence-for-sentence; explanation follows in second part. Leading phrases in square brackets identify the "point" of what follows and are not necessarily intended to be part of the text itself).

  1. [What is a secondary source]
    A Secondary source is a "work that interprets or analyzes" a previously described event or phenomena. "It is generally at least one step removed from the [previously described] event. Examples include scholarly or popular books and articles, reference books, and textbooks."[3]
  2. [Why a secondary source is important to NOR]
    These secondary sources are always preferable because they put the subject that they are being cited for into perspective, both in relation to analysis of events and phenomena, and also in relation to previous discussions of the subject. Were it not for the breath/perspective provided by secondary sources, an article could only contain literal but disjoint statements, and the suggestion of relationship between them would constitute original research.
  3. [Why a secondary source gives N/WEIGHT/RS]
    Use of secondary sources is also crucial to establish the weight of a statement: An evaluative (either explicit or implicit) statementhere I'm speaking of something citing a PS, but am not explicitly using the term 'PS' for obvious reasons that is not (also) confirmed by any secondary source is an evaluation that does not merit inclusion in an encyclopedic article, and may be removed on sight.
  4. [Why a secondary source is important to NPOV]
    Secondary sources are essential to maintaining a neutral point of view, which is another cornerstone of Wikipedia policy. For one, a source that provides perspective cannot be "balanced" with a source that does not. For another, the representation of various points of view - which are inherently evaluative - can only occur through concomitant use of secondary sources.

Explanations, by way of close-reading of the original (point numbers refer to sentence # in the original):

  1. The rewrite as point #1 above because, as Carl points out, the quotation refers specifically to "historical event[s] or phenomen[a]," and as such doesn't directly back up the generalization that it is being quoted for. However, that problem can be avoided by putting the quotation inline and replacing "historical" with a paraphrase to express what you understand by it. (Inline also because there is no reason to both paraphrase and quote a simple sentence such as this).
    For reasons mentioned in the next point, I don't think the examples of the quotation are a good idea, but I've left them in because they are part of the quotation. To replace these with other examples would itself be OR :p. Such is the danger of using a quotation that doesn't match perfectly.
  2. I've coalesced sentence #2, #5 and #6 since all three describe what intrinsic value secondary sources have in articles. I've then used the three to bring the relationship to NOR into focus, which in the original isn't explicit at all.
    In the paraphrase, I've avoid explicit examples ("university-level textbooks") because that is a can of worms that we do not want to open again. (I can almost already hear '"university-level textbook" in China is not an authority on Mao'). We should explicitly identify the feature that makes a source ideal, not identify features by way of example, which is always going to be subject to (mis)interpretation.
  3. ... and ...
  4. Both Care should be taken to avoid undue weight and ensure the information cited is used in context.
    and Secondary sources may be biased by the views of its author(s) and/or publisher, so caution is required to preserve the neutral point of view
    are not issues specific to secondary sources, and should be dealt with in the general discussion of sources (which they already are).
    Both issues are re-addressed with a focus specifically on the "analytical" qualities of secondary sources in my suggestions as points #3 and #4
  5. ... and ...
  6. folded into my point #2 above.

Also, bullets/numbers should perhaps be retained so that people citing policy can refer to the specific points by number. Or, perhaps during the draft phase put a little (#n) before each sentence.
-- Fullstop (talk) 19:56, 13 December 2007 (UTC)

"Use of secondary sources is also crucial to establish the weight of a statement: An evaluative (either explicit or implicit) statementhere I'm speaking of something citing a PS, but am not explicitly using the term 'PS' for obvious reasons that is not (also) confirmed by any secondary source is an evaluation that does not merit inclusion in an encyclopedic article, and may be removed on sight." prohibits using any evaluative statements from the first scholarly publication on a topic until the statement is backed up in a secondary source. This is unacceptable. --Gerry Ashton (talk) 20:13, 13 December 2007 (UTC)
wow. thats helpful. Why not try rephrase it as you would like to have it?
While I don't disagree with your view of what constitutes a P source, but thats not the cookie thats being crumbled here. Per NOR, your "first scholarly publication on a topic" is a secondary source, not a primary one. This is because:
a) it is academic (since p and s are mutually exclusive, and p = non-academic, academic must be s)
b) it is analytical (since the defining characteristic of s is "interpretive or analytical")
So, chill. Your source is "acceptable" after all. :) -- Fullstop (talk) 20:49, 13 December 2007 (UTC)
Your point #1 is completely irrelevant as the citation was changed to better reflect the broader usage a few hours before your comments. Another problem is the absolute language used, which is obviously not going to attain consensus. While I appreciate your effort, your suggestions and analysis are problematic in a number of ways. Vassyana (talk) 21:12, 13 December 2007 (UTC)
The passage I find unacceptable includes the sentence "A Secondary source is a 'work that interprets or analyzes' a previously described event or phenomena". So the first scholarly publication on a topic is a primary source by this definition (unless the event or phenomena has been described in an unscholarly publication first). Therefore my analysis was correct and the prohibition on evaluative statements from primary sources is unacceptable.
As for rephrasing it as I would like it goes, I prefer the existing policy to the passage we are discussing (which is not to say the existing policy is optimum). --Gerry Ashton (talk) 21:10, 13 December 2007 (UTC)

Unless someone comes up with a problem I have overlooked, I would agree to Vassyana editing into the policy the above blockquoted text, but without the last two sentences (as Kenosis suggests) and then start a discussion on if the last two should be used to replace or complement Tertiary sources. (I think it is much better to introduce focused changes and not try to change too much in one go.) I also agree with Gerry and could add another couple of problems I see with Fullstop's suggestions. --Philip Baird Shearer (talk) 20:53, 13 December 2007 (UTC)

so, lets hear them! :) Since what I wrote is based on my understanding of what Vassyana wrote, then by identifying the problems with my suggestion, we can perhaps identify problems with Vassyana's. Understand what I mean?
Also, if Gerry's objection is valid (i.e. that the "first scholarly publication on a topic" is also a primary source), then not only is Vassyana's definition screwed, so is NORs. What Gerry notes is one of the fundamental objections to the PSTS section.
-- Fullstop (talk) 21:08, 13 December 2007 (UTC)
Unless I'm completely misunderstanding Gerry (he's free to correct me), his point is that your suggestion would define them thusly, not that they are such a source. Also, I do not believe criticisms of your suggestions would provide light on my proposal. For example, I do not use absolute language such as "always", nor do I even imply that information should be "removed on sight", as examples. Vassyana (talk) 21:16, 13 December 2007 (UTC)
I refuse to decide whether the first scholarly publication is a primary source or not. I think the definitions in various fields of scholarship are too entrenched for Wikipedia to adopt its own definition, just as the U.S. transportation authorities gave up on trying to use the proper meaning of "inflammable" and now use "flammable" instead. Wikipedia policy should be framed so that it does not matter whether the first scholarly publication is a primary or secondary source. --Gerry Ashton (talk) 21:25, 13 December 2007 (UTC)
Makes sense. The law of primacy is a motherlover. ;o) Vassyana (talk) 21:30, 13 December 2007 (UTC)
I completely agree with Gerry Ashton's comment. — Carl (CBM · talk) 22:27, 13 December 2007 (UTC)
Indeed, "the definitions in various fields of scholarship are too entrenched for Wikipedia to adopt its own definition" is absolutely right. So where does that leave the PSTS section, or for that matter this talk section? :) -- Fullstop (talk) 22:45, 13 December 2007 (UTC)
While it's no secret that I also agree (I pushed for a caveat about varying definitions and support the idea of a terminology change), that's a whole can of worms unto itself. I'd recommend continuing to work on reviewing and revising the section as agreeable. Doing away with the primary/secondary distinction, even if it's just to use other terms, is going to highly contentious and require it's own collector's volume of discussion. ;o) Vassyana (talk) 23:48, 13 December 2007 (UTC)
Yep, regardless of whether we like it or not, since "secondary" is under discussion, we might as well make the most of it.
Issue 1: the secondary bit as it stands now needs something to relate to NOR.
Issue 2: if primary could (might) be defined as being what secondary isn't, then by making the secondary def independent of the primary def we might subsequently avoid the hedging/hair-splitting on what primary is.
Or, to apply the US-DOT (in)flammable analogy: when "secondary sources" is adequately well defined, all references to "primary sources" can be replaced with "non-secondary sources."
With that, its possible to have the cake and eat it too. As far as "primary" goes, there would be no (need for) terminology change because there would no terminology to change.
Get it? :) -- Fullstop (talk) 00:46, 14 December 2007 (UTC)
I suggest that Fullstop is assuming the conclusion. Fullstop does not want to let evaluative statements that occur only in primary sources into articles, thus, any definition of primary source must exclude any desireable source of evaluative statements. The existing NOR definition of primary source could easily be read as including the first scholarly publication on a topic, but it is not screwed, because the existing NOR policy does not prohibit using evaluative statements from primary sources.
The definition of a source and what one may do with the source are thoroughly intertwined. I almost wish the Wikipedia edit software could be made so that when an editor changed a single character in a source definition, the policy about what one could do with sources would be automatically blanked, and the editor would have to rewrite that part of the policy. --Gerry Ashton (talk) 21:20, 13 December 2007 (UTC)
aaah! I see where you're coming from now. No, actually the point wasn't to 'not let evaluative statements that occur only in primary sources into articles'. Instead, the point was that secondary sources contribute to the weight of notability of a subject, while primary sources don't. The corollary being what N already expresses, which is that a subject that has no secondary sources is non-notable.
In other words: If an article that has no secondary sources fails notability, then: This implies that the sources that does have are not secondary. (whether these non-secondary sources are evaluative or not is moot).
-- Fullstop (talk) 22:40, 13 December 2007 (UTC)
ps: My whole point of not even mentioning primary sources is the same as what you expressed with your flammable/inflammable analogy: Its not necessary to walk the tightrope with one if you can properly frame the other.
There is a vast difference between a topic being notable enough for Wikipedia to have an article about it, and a statement being notable enough to be included in Wikipedia. A topic with no secondary sources is unlikely to be notable enough to have a Wikipedia article, but a statement from a primary source may very well be worth adding to an article even though it is never mentioned in a secondary source. Then again, some statements that propose controversial or unlikely hypotheses may not be worth mentioning unless they are mentioned in secondary sources; it depends on the nature of the statement. --Gerry Ashton (talk) 23:26, 13 December 2007 (UTC)
Gerry, you're preaching to the choir. Read what I wrote in the context of the suggestion, which is itself a response to something else. -- Fullstop (talk) 23:38, 13 December 2007 (UTC)
I think it's close, but maybe melding best parts of this and section above. "Secondary sources are references that draw on research and other references to make interpretive, analytical, or synthetic claims, or create a general overview.[1][2] Secondary sources are usually preferable because they generally provide analysis, offer a more independent view and provide a broader context for the subject. Where interpretive claims, analysis, or synthetic claims about primary sources are included in Wikipedia articles, use secondary sources rather than original analysis by Wikipedia editors. Those views can reflect the biases of their author or publisher, so caution is required etcetera." Professor marginalia (talk) 01:43, 14 December 2007 (UTC)
A very sensible and excellent suggestion. I've added the phrase as a closing statement, as it seems very appropriate to close on that note, considering the policy context. Cheers! Vassyana (talk) 05:21, 14 December 2007 (UTC)
Just a general comment. I'm pretty impressed by the civility and teamwork from people on both side of the issue in this section the last couple of days. It's a marked difference from how things were before. The litmus test though will be to see what happens after changes are physically made to the policy. wbfergus Talk 12:50, 14 December 2007 (UTC)


Thanks for your kind words above, Vassyana - I'll try to live up to them! Re-reading after a night's sleep has brought a few more points to my mind:

  • I don't think the average editor will understand what "synthetic claims" are. They sound like they are made out of cheap plastic; or worse, something that's completely made-up. I'm not sure that the mention of "works of synthesis" in footnote2 is enough to explain. Would a wl to synthesis help, or wikt:synthesis? Or a separate explanatory footnote? I can't yet think of a reasonable rewording.
  • Can I again suggest reordering sentence 2 thus: "Secondary sources are usually preferable because they generally provide analysis and a broader context for the subject, and usually offer a more independent view. However, they can be biased..." That's saving you a "provide" and putting the bit that the next sentence is going to qualify at the end so it flows neatly into it.
  • The new last sentence uses the word "primary" which I think you're trying to avoid. For consistency, should it also use the same phrasing "interpretive, analytical…" as the first?
  • That last sentence also suggests to me that the "original analysis by WP editors" already exists, like the secondary sources do. I don't think that's the intention - it's really trying to say "don't do it yourself" isn't it?

Hope this helps,  —SMALLJIM  15:57, 14 December 2007 (UTC)

I replaced "synthetic" with "synthesized" to help avoid the "artificial" meaning (no pun) of synthetic(s). I reordered the sentence per your suggestion. I replaced that phrase in the trailing sentence with "using your own analysis". Vassyana (talk) 20:42, 14 December 2007 (UTC)

I do not think that "Where interpretive claims, analysis, or synthetic claims about primary sources are included in Wikipedia articles, use secondary sources rather than using your own analysis." adds anything that is not already in Primary Sources, so although it is not a show stopper for me, I think as it makes this new paragraph slightly more verbose and less targeted than it needs to be. --Philip Baird Shearer (talk) 17:20, 15 December 2007 (UTC)

Hi, I did not realizes there had been a major discussion prior to Vassyana’s edit. I hit the link that the edit summery referred to, but it goes to WP:NOT, rather than here. OK, having said that, what does “*Secondary sources are references that draw on research and other references” bring to the table that “*Secondary sources draw on primary sources” does not? I see no mention of primary sources, and under the new wording, I can imagine a LOT of circular referencing (Jones quotes Smith, Smith quotes Green, Green quotes Jones) going on. It’s the “other references” part that bothers me. I have not read the talk page yet, or not enough to understand what you are tying to get at here. Brimba (talk) 04:19, 16 December 2007 (UTC)
Ugh, evil QWERTY typos! ;) Essentially, it comes down to the fact that secondary sources are usually not exclusively based on primary sources. In some fields, especially within humanities and mathematics, it's not at all uncommon for there to be a heavy reliance on other secondary sources. The wording may indeed be clumsy and if you have some alternate suggestions they'd be welcomed. Also, for the trailing sentence of your edit, please check review the wording immediately below, which was left out for further comment. Vassyana (talk) 16:25, 16 December 2007 (UTC)

Implemented

I've added the section except the final statement to the live policy. I think it would be good to get a little more feedback about it before adding it to the live page. As a reminder, that closing sentence is:

Where interpretive claims, analysis, or synthetic claims about primary sources are included in Wikipedia articles, use secondary sources rather than using your own analysis.

What do people think of adding that as a closing statement for the secondary sources paragraph? Vassyana (talk) 23:01, 15 December 2007 (UTC)

Based on what's been done to your text since you made it "live", Vassyana, what I think is that you're obviously wasting your time, talents and insights trying to get anything changed here. I have no doubt you're feeling extremely frustrated again. I wonder if writing an essay would be a more profitable avenue to explore? Some essays are widely read and cited, WP:BEANS for instance. Maybe you could build up a wider consensus via that route. Hope this helps,  —SMALLJIM  11:50, 16 December 2007 (UTC)
I walked away in frustration already. I'm not doing so at this juncture. I'm not going to let basic consensus building and discussion be brushed away with frustration and "bold" editing. Vassyana (talk) 13:01, 16 December 2007 (UTC)
Wow, you might be right. I received a warning exhorting to use the talk page and respect consensus for ... enforcing actively discussed changes.[6] This from a person who claims his views aren't views, but Foundation level policy,[7] who hasn't bothered to join in the discussion of the changes or bothered to propose his own changes.[8] It's sickening and tiresome. Vassyana (talk) 17:05, 16 December 2007 (UTC)

Journal articles

Hi there, this policy is pretty confusing. Are journal articles, such as PMID 18052346 primary or secondary sources? I know the data in the article (such as the numbers in the figures and tables) are primary, but is the article as a whole a secondary source? This question isn't about changing the policy (I know better than to try!) but this needs clarification. Tim Vickers (talk) 16:40, 14 December 2007 (UTC)

Does it matter? Are you making any claims, statements, etc. that aren't already in print? If they are printed, verifiable an reliable, what difference does it really make if you or other editors are not making any claims or conclusions about the work? Either way, primary, secondary or tertiary, the statements should be sourced, so that any reasonable person can look at the Wikipeadia article and the reference and see that they say the same thing in the same context. If so, it's not original. wbfergus Talk 16:50, 14 December 2007 (UTC)

It matters because people are starting to argue that if an article does not have "secondary" sources it fails notability. Even if the article contains several journal articles about the subject as references. See for example User_talk:ProteinBoxBot#Notability - Tim Vickers (talk) 16:53, 14 December 2007 (UTC)

Tim, you're arguing on that page that journal articles are secondary sources. They're not, if they're written by the people directing involved in a piece of research. A primary source is written by someone very close to the issue at hand, directly involved in it, or an observer of some kind. The reason we require secondary sources (written by uninvolved people) is to check that the issue is really notable i.e. that people not involved find it as interesting as people who are involved. SlimVirgin (talk)(contribs) 17:36, 14 December 2007 (UTC)
That's exactly what has happened here, though in a roundabout way. The NIH has decided that the subject is notable, hence the inclusion of the data through their website. They don't include every piece of 'health' related literature, they actually make their own decisions on whether something is notable enough to warrant inclusion. They themselves have therefore become a secondary source, and notablity is met.
Okay, here's my opinion.
  1. The user there has created his own definition of notability, that very much appears quite different from WP:NOTE.
  2. It appears in the context you are using them, they would be secondary sources. Some of the data may be primary source, but if the journal itself is one of multiple sources used for the article, then there is a good chance that it would meet the notability requirements.
Again, these are my opinions, but that's how I would see the (non-)issue there. wbfergus Talk 17:08, 14 December 2007 (UTC)
Further comment. I see that the site you reference is NIH, a government web site, that is providing the article from another publication. Being in the government, I can say that rarely does the government (at least within my 'Department of XYZ') even use primary source data, unless generated in house first. Next, if the government does publish something, especially in the scientific field, it does go through a rigorous peer-review process (in my agency, around 2-3 years) before getting published. So, I would have a fairly high degree of confidence that articles listed on that site alone would meet the notability criteria by any definition. wbfergus Talk 17:15, 14 December 2007 (UTC)
Thanks, that pretty much matches my interpretation of what a "secondary" source is. Is this a controversial definition? I wish the policy was a bit clearer, since we are obviously having difficulty understanding it unambiguously. Tim Vickers (talk) 17:28, 14 December 2007 (UTC)
A journal article outlining a study written by the people involved in the study would be a primary source of information about that study. SlimVirgin (talk)(contribs) 17:33, 14 December 2007 (UTC)
Maybe. Citing the NIH website itself makes the NIH a secondary source that uses a (maybe) primary source. NIH is a highly regarded source of information within their area of expertise. If NIH has decided to allow the inclusion of the article on their website, they must have already made the decision that the subject is notable, and that is the entire breadth of this 'controversy'. wbfergus Talk 17:41, 14 December 2007 (UTC)
Journal articles make generalizations and interpretive, analytical, or synthetic claims, this is what the discussion and conclusion sections of an article are for. If journal articles themselves (not the raw data they contain in the results section) are indeed primary, the definition of secondary used here is incorrect and would need to be changed. Tim Vickers (talk) 17:50, 14 December 2007 (UTC)
Agreed. It just goes to help highlight how problemmatic these Wikipedia specific definitions are, even if they are loosely based on the Library sciences. There are many other fields other than just library sciences with different definitions. It further complicated that these terms and definitions are included in a policy, rather than a guideline that is more open to interpretation. Polices shouldn't be open to interpretation. wbfergus Talk 17:57, 14 December 2007 (UTC)
Two issues are getting mixed together here. Secondary sources are needed to establish notability of a research topic or finding. Primary sources, on the other hand, are better for showing what the finding actually was. Raymond Arritt (talk) 18:00, 14 December 2007 (UTC)
For the purposes of WP, peer-reviewed articles that analyze data, draw conclusions, or make interpretations are considered secondary sources. The actual data gathered in experiments is considered a primary source. — Carl (CBM · talk) 18:02, 14 December 2007 (UTC)
Carl, who says this? ... Kenosis (talk) 00:33, 15 December 2007 (UTC)
Certainly, that's how I've generally seen peer-reviewed scientific literature treated (like secondary sources). Vassyana (talk) 07:05, 15 December 2007 (UTC)
My brain hurts. Raymond Arritt (talk) 18:09, 14 December 2007 (UTC)
I have the agree with CBM. Slim's broad version of "primary sources" would essentially limit many academics topics to tertiary sources for establishing notability. This is especially true in the humanities, where nearly all secondary sources (assuming tertiary sources are not counted as secondary sources) include original research of their own, which would make them primary sources under Slim's assertion. Vassyana (talk) 20:21, 14 December 2007 (UTC)
WP:PSTS has nothing to do with WP:Notability. Notability is a term used to denote whether an article is adequately notable, or if you prefer, sufficiently noteworthy, to be included in Wikipedia. ... Kenosis (talk) 00:38, 15 December 2007 (UTC)
Tim, I agree. When we wrote WP:ATT as a replacement for NOR and V, we tried to make the definitions clearer, but there was nothing but fighting so we gave up. The following is from an early version of ATT (and it could easily be expanded to discuss journal articles):

Primary sources are documents or people very close to the situation you are writing about. An eyewitness account of a traffic accident is a primary source. The Bible is a primary source. The White House's summary of a George Bush speech is a primary source. Primary sources that have been published by a reliable source may be used for the purposes of attribution in Wikipedia, but only with care, because it's easy to misuse primary sources; for example, anyone could try to use the Bible as evidence that God said homosexuality was a sin. For that reason, edits that rely on primary sources should not interpret the source material, but should simply describe it. Any interpretation of primary source material requires a secondary source.

Secondary sources are documents or people that summarize other material, usually primary source material. A journalist's account of a traffic accident is a secondary source. A theologian's account of what the Bible says is a secondary source. A New York Times account of a George Bush speech is a secondary source. Wikipedia articles should rely on reliable, published secondary sources wherever possible. This means that we publish the opinions of reliable authors, and not the opinions of Wikipedians who have read the primary source material for themselves. As a rule, we want to know what Professor Smith, the theologian, says about the Bible and homosexuality, and not what User:Smith says about it, even though both are relying on the same source material.

SlimVirgin (talk)(contribs) 18:10, 14 December 2007 (UTC)
In articles on ancient architecture, the ruins themselves are a primary source. If someone else comes along and draws (and publishes) a layout of the ruins, especially if parts are missing so the author concludes how it must have been, then that is a secondary source. wbfergus Talk 18:18, 14 December 2007 (UTC)

congratulations all for demonstrating once again how not-a-good-idea the words "primary" and "secondary" are, that it undermines other policies (here notability), and that it is less painful to invoke common sense (aka WP:IAR) than butt heads with the ambiguities/contradictions of PSTS. :) -- Fullstop (talk) 19:05, 14 December 2007 (UTC)

I think we've got this sorted now. Professor Smith, publishes a paper where he discovers that a type of bacteria has a novel gene sequence X in it and he interprets this to show the bacteria are descended from group P of bacteria, rather than group Q. Citing Professor Smith that the bacteria are descended from group P is using the paper as a secondary source - something that has analysed and interpreted data. Doing your own analysis on sequence X and saying the Professor is wrong is using the paper as a primary source. Is this correct? Tim Vickers (talk) 19:34, 14 December 2007 (UTC)
Yes, that's right. Professor Smith can analyse all he likes -- he is the primary source, and if we describe his analysis, we are still only describing. It's when we want to analyse Smith's views, or use them to make a point, that we need to find a secondary source who supports what we're saying.
The more controversial the topic, the greater the care we have to take to rely on secondary sources for any contentious point. What we're basically saying to editors is -- check that other writers agree with how you're using this material. SlimVirgin (talk)(contribs) 20:08, 14 December 2007 (UTC)
Tim, sorry, I misread your last sentence. Just to clarify -- you wrote: "Doing your own analysis on sequence X and saying the Professor is wrong is using the paper as a primary source."
Yes, but any use of the paper is the use of a primary source. Professor Smith's paper is a primary source. When we say: "Professor Smith said X," we are using the primary source correctly. We are just saying what was in the paper. What we can't do is slip it into an inappropriate context, or join it with other primary sources to make a point none of those sources has made (which would be a synthesis of the kind not allowed by NOR). That is, we have to be very careful never to go beyond what the source himself has said. If we want to go beyond it, we must find a secondary source -- an article about Professor Smith's paper -- that makes the points we want to make, or that uses Smith's paper in the ways we want to use it, if we want to use it to support a point we're making. SlimVirgin (talk)(contribs) 20:13, 14 December 2007 (UTC)
I strongly disagree with your source classification. For starters, it essentially "breaks" notability. See my response to CBM above. Additionally, this would conflict with the definition of secondary sources in the proposed replacement above. Vassyana (talk) 20:30, 14 December 2007 (UTC)
Which part do you disagree with? It's a description that is entirely in keeping with how the terms are used. Where is the proposed re-definition of secondary sources? SlimVirgin (talk)(contribs) 00:50, 15 December 2007 (UTC)
I disagree that a source is primary "if they're written by the people directing involved in a piece of research". As I said, that makes quite a large number of academic sources primary, including most humanities sources. I think that is a significantly more broad interpretation than the norm. The proposal for the secondary sources replacement is at the appropriately named Wikipedia talk:No original research#Secondary sources. Vassyana (talk) 07:04, 15 December 2007 (UTC)
Again, as above, WP:PSTS has nothing to do with WP:Notability. Notability is a term used to denote whether an article is adequately notable, or if you prefer, sufficiently noteworthy, to be included in Wikipedia. ... Kenosis (talk) 00:46, 15 December 2007 (UTC)
Notability's usage of "secondary sources" is dependent on PSTS. It's even directly linked in N's "in a nutshell". Vassyana (talk) 07:04, 15 December 2007 (UTC)
Ah! I'm glad it's not just me. I asked Kenosis about this before.[9] His reply[10] did not help. I'd be very grateful if you could explain, Kenosis, because I honestly do not understand what you mean.  —SMALLJIM  10:10, 15 December 2007 (UTC)
First off, if it's that confusing, the link from WP:Notability to WP:PSTS should be removed, and if need be, replaced with a self-standing definition. Note that WP:Notability also links to WP:Reliable sources. I find it difficult to believe it's that confusing to editors who are as articulate as the two who just stated this. Notability is, as would be apparent at this stage of reading into WP policy by Vassyana and Smalljim, a separate standard having to do with setting a principle to establish worthiness for inclusion in Wikipedia. In other words, pick a topic-- just about any topic: if it hasn't been discussed by secondary sources, it ain't notable for purposes of discussing WP:Notability. Here, in WP:PSTS, secondary sources are required for analysis and synthesis of such primary source. Other than to poke holes or argue that the "vox populi" won't have a clue, what's the problem with this overlap where the editors of WP:Notability have chosen to draw on the present definition of secondary sources offered in WP:NOR? If the recently proposed text for "secondary sources" is inserted (the one with all the extra commentary about undue weight and NPOV and such, maybe it would be sensible to go over to WP:Notability and remove that link, or perhaps even to link cross-namespace to the article on secondary sources. ... Kenosis (talk) 14:12, 15 December 2007 (UTC)
Thank you for the response, Kenosis. It was most illuminating.  —SMALLJIM  16:42, 15 December 2007 (UTC)
Kenosis, that approach may work, but wouldn't it ("replaced with a self-standing definition") eventually be yet another point of confusion for Wikipeia editors, where another definition of 'source-typing' would be allowed to morph into something different from other usages on Wikipedia? Wouldn't having one common defintion somewhere on Wikipedia be much more consistent for all of Wikipedia and all Wikipedians, even if it did remain here? All articles, guidelines, policies, etc. link to the same exact definitions to avoid any confusion or subsequent morphing factors. Otherwise, the confusion factor would explode if every article, guideline or policy was 'allowed' to create their own unique definition of something that has so much carry-over into other areas. Thanks. wbfergus Talk 14:25, 15 December 2007 (UTC)
The standard at WP:Notability is essentially reduced in the nutshell, by those who participated in that guideline, to (1) whether there is a sufficient number of secondary sources written about the topic, beyond the primary source of that topic, to deem it adequately noteworthy to have an article about it, and (2) whether such secondary sources are deemed sufficiently reliable to have an article about it. Thus, the concept of primary and secondary sources, in conjunction with the concept of reliable sources, is used in that analysis. Is this potentially confusing in and of itself? I suppose it can be. Add another use of "secondary sources", is it potentially more confusing? Sure. Add annother use of "reliability" (in WP:V, is it potentially more confusing? Absolutely. Add WP:NPOV, is it yet more confusing? I should think so. Add the concept of WP:Undue weight--more confusing? Yep. Add the concept of WP:Consensus, is it more confusing? Ah yes. Add the various other ins-and-outs of the policy pages, and is it a set of policies that no one can necessarily feel secure that every edit they make can be stricly defended on a policy basis, as in "hey, I'm a followin' the law here!"? Yes. It's editorial policy in an open-source encyclopedia that anyone can edit, which by way of being an editorial policy necessarily includes a range of discretion by the editor and the possiblity that others may disagree. Is there a cut-and-dried solution to the problem that disagreements will occur? No-- absolutely not. ... Kenosis (talk) 14:41, 15 December 2007 (UTC)
Okay, I see where you are coming from and agree (more or less) with what you said. But, I think maybe I didn't articulate my question well enough. If this policy haas one definition of 'secondary sources', and the article secondary sources had a different definition, and the notability had yet it's own (eventual) definition, they could all easily take off on different paths, using different fields of study as the basis for their definitions. This (NOR's) definition could be based on library science and/or histriography, 'Secondary sources' based on Engineering, and 'Notability' based upon science (whatever those definitions turn out to be. Rather than creating a 'sphaghetti code' effect of making users follow a circuitous route back and forth, jumping from one policy, guideline, or article to another trying to get to the definition they feel meets their needs, wouldn't it be far more beneficial to all Wikipedia users to just have one 'place' to go to, where the varying definitions could be aggregated into one cohesive collection, even if broken out by 'subject matter'? For instance, separate sections for what constitutes PSTS for each field of study, (very) roughly similar to the 'Various Examples' in the Sandbox here? There would of course be varying definitions, but it would be easier to quickly see which definition applied to which 'case', rather than jumping around through 15 different articles and still possibly missing the correct one. Thansk again. wbfergus Talk 15:01, 15 December 2007 (UTC)
You're welcome. ... Kenosis (talk) 15:10, 15 December 2007 (UTC)
It's a nice idea, wbfergus, but I think it would be unworkably complex. I've said elsewhere that the only way we could develop uncontentious policy definitions of what we want "primary" and "secondary" sources to mean would be to invent our own new terms for them. (I think I suggested Wiki Type A, and Wiki Type B sources). But that won't happen, for obvious reasons.  —SMALLJIM  15:45, 15 December 2007 (UTC)
It would be pretty strange to be discouraged from citing the conclusions of a peer-reviewed article in Nature or Science, and encouraged instead to rely on reports on the research in sources such as newspapers, which will be much less accurate and reliable. Tim Vickers (talk) 22:54, 14 December 2007 (UTC)
Ouch. Yes it would be pretty strange if that's what the policy said, but it's not what the policy says. The policy gives fairly specific limits to how to use primary sources, whatever the consensus on a given topic may judge the primary sources to be. (Journal articles are not necessarily primary by merit of being journal articles, nor are they necessarily secondary by merit, as someone said earlier, of being peer reviewed). It's not that much of a problem in comparison to the overall effort expended in researching and editing a topic where there are points of potential confusion, or sources that contradict one another, or economically or politically influenced disputes among the available sources. Numerous articles have successfully used the PSTS approach in balancing their sourcing so that primary sources are quoted and attributed where relevant or closely paraphrased in such a way that reasonably educated persons without specialist knowledge can go check it and arrive at a consensus that the source says what the WP article says it says. Where interpretation is involved, you're going to need secondary sources. If, as Tim Vickers mentions, the only options are (1) an article in Nature introducing a new topic or theory, and (2) the newspaper reports, then the newspaper reports are likely to contain quotes of what other competent scientists and researchers say about the new topic or theory in the Nature article. If it's been in the newspapers, then other scientists and researchers will publish secondary sources about the particular topic or theory shortly enough (these days the first responses often are out on the web before the hard copy even actually hits the news stands). Thus the choices are not limited to the "primary source(s)" (whatever the local consensus at an article deems this to be w.r.t. that particular topic or part of an article) and the unreliable pop science and mass-market journalism. ... Kenosis (talk) 01:28, 15 December 2007 (UTC)
Only if you're going to analyse, Tim. If you're just reporting, the primary sources are fine. And for analysis, it doesn't have to be a newspaper. Nature is fine as a secondary source too. SlimVirgin (talk)(contribs) 00:50, 15 December 2007 (UTC)
I suspect the issue isn't original research, it's notability. If an article in Science by the authors who performed experiments is a primary source, and the topic isn't mentioned in some secondary source, it isn't notable, even though in reality the topic of every single article in any 21st century copy of Science is more notable than much of what appears in Wikipedia. --Gerry Ashton (talk) 01:19, 15 December 2007 (UTC)
Brief digression: I wonder if the following articles are WP:Notable? 101st meridian west, 102nd meridian west, 103rd meridian west, etc. etc. ... Kenosis (talk) 01:52, 15 December 2007 (UTC)
This is very perplexing. We have a policy, but there is no consensus on what it means. Tim Vickers (talk) 01:54, 15 December 2007 (UTC)
I don't see your point. In what way is that different from any other Wikipedia policy? Raymond Arritt (talk) 01:55, 15 December 2007 (UTC)
:) LOL Tim Vickers (talk) 02:03, 15 December 2007 (UTC)

Weighing in

The way to solve problems about anything on Wikipedia is to use sources. In this case, we need sources that tell us what primary, secondary and tertiary sources are in the scientific literature. I'm just going to list a large number of links to reputable websites without saying what each one is, in the hope that the volume of evidence will demonstrate that a primary source in science is a journal article written by the experimenters that reports on the results of an experiment, a secondary source is a review article or possibly a meta-analysis, and a tertiary source is something like a textbook, an encyclopedia, or a reporting in the popular press. I'm going to do this blind; I have no idea what the sources will say. Also, I haven't looked in a while, but I'll bet that Wikipedia's own articles on sources, when they mention the meaning of the term in the scientific literature, will approximate what I have said above. I'll be back in a few with the links. AnteaterZot (talk) 05:46, 15 December 2007 (UTC)

What say you? AnteaterZot (talk) 06:03, 15 December 2007 (UTC)
There is enough variety in source typing definitions that for everything but the extremes you can find a reliably sourced definition to suit whatever stance you want to take. Meta-analysis and review level literature is usually considered a tertiary source (like textbooks and encyclopedias) under most definitions. Literature documenting original research is commonly referred to as primary sources within the scientific fields. However, the usage and meaning is a bit different than how the source categories are used in Wikipedia. In the Wikipedia sense of the terms, peer-reviewed articles are certainly treated as secondary sources (used for analysis, to establish notability, etc). Vassyana (talk) 07:01, 15 December 2007 (UTC)
Can you provide any sources that say that about the scientific literature? AnteaterZot (talk) 07:25, 15 December 2007 (UTC)
Not sure if any of the links or definitions on this page will help or not, WP:No original research/Sandbox/Various examples. First, let me say that if you find any specific definitions of 'source-types' as they apply to the science field, please add them to that page, for the benefit of others later on.
Next, let me briefly try to recap my earlier statements in regards to this current 'conflict resolution case' between you and TimVickers. This dispute seems solely about whether or not an article is 'notable' enough for some programmatic 'bot to tag articles semi-accurately. So, in a small way this is about what is a primary source vs. a secondary source, but mainly this is about notability. So, quoting the nutshell from the notabilty page, "A topic is presumed to be notable if it has received significant coverage in reliable secondary sources that are independent of the subject". So, in this particular case, if the citation is made to the NIH website hosting the article, then NIH is clearly a secondary source, they did not conduct the research. Next, NIH is definately considered a reliable source within their area of expertise. Since they are more-or-less 'hosting' the article, somebody there at NIH has already made the decision that within their field (health), the article is notable enough to make available to others, as it is either pertinent to other ongoing work, ties in with previous work, or is one of the first in a new series of work. Either way, it is notable to them (and they probably know about this field than most Wikipedians). Also, though I don't know the sepcifics of how NIH does it, in the government agency I work in, in order for an 'outside' publication to be on one of our servers for serving to the public, some sort of group of subject matter experts has to have given the okay after they reviewed it. So, in accordance with WP:AGF, I would be sorely tempted to say the same process is in use there. They have so much work, so few people, and such limited dollars to spend each year, they made the decision that the subject (article) is accurate enough by their standards to make available, as it offers additional information to something (probably some specific field) that they are studying or working on. So, it comes through a reliable secondary source, which had it's own peer-review process to decide if it was accurate and pertinent to their work.
As a government worker myself in another science agency, I know how long it takes for the publication of scientific studies in-house (2-3 years just for the peer-reviews and editorial oversights) after the author(s) decide the publication is ready. NIH is probably the same way, so their 'offering' of the article is more than likely their way of making information they feel is notable and pertinent until such time as they can make their own 'publication' available. I would suggest (without knowing what your 'bot does or is planned to do) that if it finds articles with less than two references, it simply tags the artilce with the {{expert-subject}} tag. wbfergus Talk 11:44, 15 December 2007 (UTC)
I understand and respect that there's a tension between WP:PSTS and WP:NOR on the one hand, and the request made by some users for expert help in certain articles, for instance by placing the {{Template:Expert-subject}} tag. To this I can only say that when a user becomes involved in a WP article and says "hey, I'm an expert on this topic", if that user is indeed an expert, that user should be able to much more readily point the way to the reliable secondary and/or tertiary sources on that topic, rather than using WP as an opportunity to publish original research. Use of primary sources about particular theories, concepts, literary styles, technical approaches, etc., should be limited to verbatim representations attributed to given primary sources in such a way that any generally educated person can double check the primary source and conclude that it says what the WP article says it says. Analysis, synthesis, comparison, explanation or integration vis-a-vis other theories, concepts, literary styles, technical approaches, etc., should use secondary sources, ideally citing representative secondary sources. Any expert in a given topic area who is willing to participate in helping the project by improving such an article should be able to accomplish this. Upon being presented with this policy, it's only a problem if the expert's goal is to publish original research in Wikipedia. ... Kenosis (talk) 15:01, 15 December 2007 (UTC)
In mathematics at least, the place that commentary on "analysis, synthesis, comparison, explanation or integration vis-a-vis other theories, concepts" is often found is in peer-reviewed journal articles. That's one reason such articles are treated as secondary sources on WP. Original research in this context would be the promotion of theores that have not appeared in peer reviewed papers. — Carl (CBM · talk) 15:11, 15 December 2007 (UTC)
I don't know who treats them as necessarily secondary sources, but I'll immediately point to Watson and Crick's article in Nature about DNA. Is that a primary source? You bet it is. Watson and Crick's 1953 article in Nature is a primary source for the concept that DNA molecules have the form of a double helix. And there are an immensely large number of additional examples where the particular principle under discussion was first introduced in a peer-reviewed journal (I will leave aside for the moment the residual debate about the extent of peer review in Watson and Crick's case). Moreover, to the extent that a peer-reviewed article is documenting follow-up on a concept, theory, topic, or particular type of analysis, but the new synthesis or analysis is sufficiently original to be given a name of its own, it too is a primary source for that particular brand of approach to the already existing topic. Why is it a primary source in such a case? Because it's the original source of the concept, type of analysis, or particular variation that was given a particular name for that concept, type of analysis or variation. An example might be modern evolutionary synthesis, which has a couple of primary sources for the concept itself, despite that it was built upon earlier primary sources such as Darwin's Origin of Species. And, to follow this one example a bit further, Origin of Species is not the primary source for every little thing within its covers either, but it is a primary source for the theory of evolution by natural selection. And each new variation that is significant enough to be given a name has at least one primary source, which quite commonly is first published in a peer-reviewed journal. As is stated in the policy page under WP:PSTS it depends upon what is being talked about in the WP article. ... Kenosis (talk) 15:25, 15 December 2007 (UTC)
(←) 'm afraid you're confusing the "first appearance" of a scientific theory with the primary/secondary source distinction. It is perfectly possible for a secondary source to introduce a theory, and indeed this is the way that most theories are going to be introduced. The point of the primary source distinction in historical analysis is to separate "eyewitness testimony" from "distanced analysis". In scientific papers. the data collected corresponds to the eyewitness testimony, but the introduction of abstract theories does not. This is one reason the entire distinction is less useful for scientific papers than for historical analysis.
Nevertheless, this is not really a significant issue. The actual problem I think you are trying to address is undue weight, not original research. By plain language, a theory that has appeared in a peer reviewed journal cannot be called 'original research' if restated on Wikipedia. — Carl (CBM · talk) 15:51, 15 December 2007 (UTC)
It is true that according to current WP policy the use of material from a peer-reviewed journal article in a simple restatement of relevant passages from such a source is not original research. I must disagree, though, that the idea that all peer-reviewed journal articles are secondary sources is not a significant issue, because it's been one of the argumentae ad absurdum used in opposition to PSTS itself. While documented data sets are primary sources, so too the first publication of a theory that is given a name of its own is always to be considered a primary source, because it's the original published source of the particular theory. If one wants to add to that the fact that lab notes and other documentation and such are also primary sources, fine. But the notion that for purposes of Wikipedia editing the first published source is not a primary source? I surely dunno about this interpretation. The original published source of anything is inherently a primary source. We're not historiographers, not data miners, and not professional analysts, but rather we're editors of a general encyclopedia that has a proscription against original research. Thus the idea of relegating the concept of "primary" to the backwaters of archived lab notes, fossils in caves, and the like, is contrary to the spirit of WP:PSTS, in my estimation. I still would like to know where all these editors have "generally accepted this idea of what a primary source is in such a way as to automatically regard peer-reviewed journal articles as secondary sources. I think it's putting the cart before the horse-- the horse here is "what is the particular thing being talked about", not what forum the source was published in. If the thing being talked about-- the particular topic-- is a theory, its first published appearance(s) is or are the primary-source material. That's why, beyond a set of representative examples given on the policy page, the decisions are made locally in the context of particular articles, not from WP:Policy Central. ... Kenosis (talk) 16:29, 15 December 2007 (UTC) ... :(My offhand sense is that the idea that peer-reviewed journal articles are secondary sources may involve some degree of confusion with WP:V#Reliable sources and WP:Reliable sources, at least one of which gives peer-reviewed journals as an example of a reliable source.) ... Kenosis (talk) 16:38, 15 December 2007 (UTC)
I think that a scientific review article is always a secondary source. An "original" article is partly secondary source (as much it summarizes previous research in Introduction and Discussion), and partly a primary source. But I would like to emphasize that annotations and other materials published in established biological databases, such as Pfam or UniProt, should be considered as reliable secondary sources according to the existing WP criteria. There is fact checking policy there, and all records have proofs as links to published journal articles.Biophys (talk) 17:24, 15 December 2007 (UTC)
In my mind there's absolutely no question about it. Review articles are inherently secondary, unless the author is using it as an opportunity to inject a particular new synthesis by the author, in which case they're primary only w.r.t. the particular new synthesis of the already-published material they're reviewing. In general, they're inherently secondary, i.e., designed to give those conversant in the particular area of research an overview of recent developments in that area of research. ... Kenosis (talk) 17:33, 15 December 2007 (UTC)
>> The original published source of anything is inherently a primary source.
By that definition, the 10 commandments are a primary source, and the old testament a secondary one. Oh, wait! The old testament is reiterated in the bible, so the bible must be a "tertiary" source. Charming. ;)
>>If the thing being talked about-- the particular topic-- is a theory, it's first published appearance(s) is or are the primary-source material.
Uh, yes, and? Does that somehow disqualify it as a "source"? If so, why? And if not, why? And in either case, how is this relevant to whether someone can screw up the representation of it on WP?
>> we're editors of a general encyclopedia that has a proscription against original research
This is perhaps the reason why you're so confused. WP doesn't have "a proscription against original research." WP has a proscription against original research by editors themselves.
-- Fullstop (talk) 17:10, 15 December 2007 (UTC)
Fullstop, based upon recent talk-page comments it seems to me you're enough of a skilled logician to know full well that just because "if A, then B", it's not ncecessarily the case that "if not-A, then not-B". Translated: "The first published source is always a primary source. But not being the first published source doesn't necessarily mean that it is not to be considered a primary source." As many have been trying to say here, including myself, it depends on the topic. If the topic is focused on historical variations on the commandments and how they were distilled to ten in the Torah and Old Testament, then the range of what's considered primary-source material will depend on what the editors discussing and editing the topic see as the original sources. If the thing being talked about is variations in the scriptural version of the ten commandments, the editors will need to cite any particular scriptural version of the ten as a primary source, and rely on secondary-source commentary from reliable sources for further analysis and commentary.
But, to be somewhat direct, it appears increasingly apparent to me that part of the objective of some recent commentators about this policy is to go as far as possible to permit original research. To the extent that this may be the case, it seems to me that removing WP:PSTS immediately gets one a large part of the way towards this objective. For my own part, I vehemently oppose such a change in existing policy. ... Kenosis (talk) 17:27, 15 December 2007 (UTC) ... Anyway, for now, gotta go. Take care, 'tll another time. ... Kenosis (talk) 17:42, 15 December 2007 (UTC)
yeah, I figured that you'd correctly point out that "not being the first published source doesn't necessarily mean that it is not to be considered a primary source." But I wanted you to say it. ;)
If there are really commentators here who wish "to permit original research," I must have missed them. All I've seen is brouhaha over PSTS. And that can't be what you're referring to since a distinction between primary and secondary contributes nothing to an understanding/explanation/prevention of original research. Removal of PSTS would get them nothing.
Sticking to the sources is sticking to the sources. Even if the shoddy PSTS were to vanish, the rest of NOR prohibits original research. And it does even without the ifs, buts, & perhapses of PSTS.
Because "the original published source of anything is inherently a primary source," original research makes an article itself a primary source, and an encyclopedia cannot be permitted to itself become a primary source. This is the only relationship and purpose that a definition of "primary sources" has to "no original research."
-- Fullstop (talk) 18:29, 15 December 2007 (UTC)
And that was the original use of "primary sources" at NOR, per this diff. Since then, the concept of PSTS has grown to cover all sorts of things well beyond the original intent. In my opinion, it has been co-opted for purposes other than preventing original research. Whether these purposes are good and pure are not of concern here, we should limit the NOR policy to defining OR as clearly as possible in order to make it easily understandable. The PSTS section just makes OR clear as mud, and causes new arguments over what is primary vs. secondary, rather than what is OR vs. NOR. Dhaluza (talk) 18:21, 16 December 2007 (UTC)
(R to Kenosis). You said, "so too the first publication of a theory that is given a name of its own is always to be considered a primary source, ..." By that logic, a contemporary book about the French revolution that advances a new theory about the revolution is a primary source, which contradicts even the definition of primary/secondary source used in historical scholarship. — Carl (CBM · talk) 19:28, 15 December 2007 (UTC)
I think that depends on whether the theory causes a paradigm shift in the discipline or whether it is just another rejected revisionist theory. In general I would not consider a "normal" paper that develops a new facet within a paradigm a primary source, but a seminal paper that develops a theory and sets a discipline off on a whole new tangent probably is a primary source. Do not ask me to judge what such a paper is in disciplines I know little about, and of course if you think that Thomas Kuhn and the secondary sources he engendered are not correct then that is a whole different debate:-) --Philip Baird Shearer (talk) 20:48, 15 December 2007 (UTC)
But once again, doesn't this show that the delineation between primary and secondary sources is complex, if not convoluted? And why is it really necessary? Can't we formulate a more specific definition of what constitutes original research, without the diversion of source typing? Dhaluza (talk) 01:54, 16 December 2007 (UTC)
Resp. to Philip. Any paper or (more likely) book that analyzes first-person accounts of the French revolution to present an interpretation of those events would be a secondary source about the revolution. It would not matter whether the book presented a completely new theory or merely extended a previous theory. In other words, the novelty of the interpretation does not make a primary source out of any contemporary book about the French revolution. Similarly, a math article that analyzes an existing mathematical concept to draw new conclusions is a secondary source about that mathematical concept. — Carl (CBM · talk) 04:31, 16 December 2007 (UTC)

Might I just suggest something - the difference between a primary source and a secondary source is relative. Relative to the subject the source covers and sometimes relative to the field it is published in. For example if I have an essay published about the latest production of Hedda Gabbler in a Theatre studies journal then there would be 2 ways to use it as a source. First the essay published in the journal is the primary source for my views, but it is a secondary source for information about the production (the production itself being the primary source for that). So if you were writing an article about Hedda Gabbler you could use that article as a secondary source and say: "X is Cailil's interpretation of the meaning of that stage direction" but if you were writing an article about me or my theories the article would be a primary source and could only be used descriptively.

What makes things difficult is the apparent difference between Hard and Soft science publications. But again the source type will be relative to the subject of the article. If I came up with a new theory of relativity and it was published in a physics journal, the journal article would be a primary source for an article about this new theory. But if that article discussed Einstein (in a way that is not fringe) then it could be used as a secondary source on Einstein's page.

Perhaps PSTS is not eloquent prose but it does help people see that "sources closest to the subject" should only be used for description not interpretation--Cailil talk 03:15, 16 December 2007 (UTC)

In reply to CBM yes and no, as Cailil points out it as depends on the focus of the Wikipedia article. If it is about the French revolution then it is as you say secondary source. But if the article is about the theory and I can think of one example Historical revisionism, and a possible Genocides in history, where such a book might be a primary source (See Historical revisionism#French attacking formations in the Napoleonic wars and Genocides in history#France). --Philip Baird Shearer (talk) 07:06, 16 December 2007 (UTC)
Even in the article about genocide,this would be a secondary source, since it analyzes primary sources to make the interpretive claim that a genocide occurred. If the book were used as an example in the article about historical revisionism, there it would be a primary source. But for an article on the French revolution or genocide, it would be a secondary source. — Carl (CBM · talk) 13:09, 16 December 2007 (UTC)

So far, I am unrefuted

So far, nobody here has come forth with a single reputable source that defines primary and secondary sources in the scientific literature any different that I did. I'll paste my sources again, below, with some annotation. Many of these are university library's pages.

  1. University of Hawaii "How does Biological literature evolve over time?"
  2. William Madison Randall Library "Identifying Primary, Secondary, and Tertiary Sources"
  3. Cal State Northridge "Types of Resources"
  4. University of Minnesota Bio-Medical Library "Primary, Secondary and Tertiary Sources in the Health Sciences"
  5. Diablo Valley College "Identifying Primary, Secondary and Tertiary Sources"
  6. Florida Institute of Technology "Primary, Secondary, and Tertiary Sources"
  7. Yale University "Lesson 3: Recognizing Primary, Secondary, and Tertiary Sources"
  8. OhioLINK, is a consortium of 86 Ohio college and university libraries, and the State Library of Ohio "Cycle of Scientific Literature"
  9. University of California Santa Barbara "Distinguishing Between Primary, Secondary and Tertiary Sources"
  10. eric.ed.gov Education Resources Information Center "Chemical Literature Exercises and Resources"
  11. And finally, a book: Writing Guidelines for Postgraduate Science Students By John Hampton.
  • I'll ask again: Does anybody have any sources about what constitutes primary and secondary literature in science to refute the sources I give above? All my sources say that in the scientific literature, a report on an experiment by the experimenters is primary, and a review article is secondary. Those of you in the humanitites, please understand that science is different. AnteaterZot (talk) 12:42, 16 December 2007 (UTC)
Please read my reply to you again. I perfectly acknowledge how science typically defines their sources. Wikipedia policy deals with all fields and therefore does not conform to one particular subject's treatment, but rather deals with the source distinctions in a way appropriate for Wikipedia. We are not publishing scientific literature here, we are creating a general-use encyclopedia. Different purposes require different standards. The footnote at the beginning of the PSTS section deals with this matter (Wikipedia:No original research# note-0) and the text explicitly states that the definitions are drawn for our own purposes here ("For the purposes of Wikipedia content policies and guidelines, primary, secondary and tertiary sources are broadly defined as follows:"). Vassyana (talk) 12:44, 16 December 2007 (UTC)
Even sticking to the definition at Wikipedia:No original research#Primary, secondary, and tertiary sources, journal articles (those that are not reviews) count as primar:y since "are very close to the origin of a particular topic." Scientists, like everybody else, put the best possible face on their work, and that's why it is necessary to have a secondary source (a review article) to make sure the topic is as important as they say it is. AnteaterZot (talk) 13:06, 16 December 2007 (UTC)
Oh, and Vassyana you put that note in there. You are arguing in bad faith. AnteaterZot (talk) 13:14, 16 December 2007 (UTC)
Hardly in bad faith at all. I did not insert the lede paragraph statement into policy. The caveat footnote was simply an observation based on months and months of talk page discussion. Its insertion was discussed and was originally part of the main text. Due to some objections over the bloat of the text, it was moved to a footnote. The other footnote was added to a proposal because other editors expressed a need for footnotes with a broader definition that better encompasses the breadth of topics on Wikipedia. I simply added two reasonable footnotes expressing the broader intent of the policy per the feedback of others. You make it sound as though I put that material into policy of my own accord, when those changes were made based on the feedback of others after building a consensus. Vassyana (talk) 13:48, 16 December 2007 (UTC)
If you say so. AnteaterZot (talk) 13:53, 16 December 2007 (UTC)
The "close to the origin" language is not particularly precise. Primary sources report data; secondary sources interpret data to make analytic and synthetic arguments. You may be thinking mostly of experimental sciences, where a paper consists mainly of data with a small amount of analysis. Papers like that should indeed be treated as primary sources. On the other hand, in other areas such as computer science and mathematics, most papers have no experimental data, and consist entirely of analytic and synthetic arguments. Such papers, by their nature, are secondary sources. — Carl (CBM · talk) 13:18, 16 December 2007 (UTC)
Thanks you for your well-reasoned response. This debate is occuring here over a project to make 10,000 stubs for genes, using a lot of scientific papers that consist of little more than reporting of data; "We found a gene, it has sequence ATGGC..., it makes a protein that looks like a protein in another species based on a computer program that looks for these things." I've been suggesting to the folks at User talk:ProteinBoxBot that they might run into trouble with notability without review articles. AnteaterZot (talk) 13:31, 16 December 2007 (UTC)
This sounds more like an issue resting on N's requirement for multiple reliable sources, with the intent of requiring enough sources to produce an encyclopedic article (in accordance with WP:NOT, from which N originally derives). Does that make sense? Vassyana (talk) 13:48, 16 December 2007 (UTC)
I guess. The editor who brought this debate here probably should have taken to to the notability talk page, do you agree? AnteaterZot (talk) 13:53, 16 December 2007 (UTC)
It is a question of notability, I agree. It is complicated by the question of how far our mission to be encyclopedic extends. For example, some claim that all towns should have articles regardless of the existence of reliable sources about them. The catchphrase used for that is "inherent notability", even though it isn't about notability so much as it is about the mission of the project. So it will require some discussion to determine which genes should have articles. WP:N or (preferably) the village pump are good locations for that discussion. — Carl (CBM · talk) 14:00, 16 December 2007 (UTC)
I agree with Carl that N's talk page and the policy village pump are the best places to address the matter. Vassyana (talk) 14:18, 16 December 2007 (UTC)
I also agree that those two venues would be the best place. However, as I remember it, the discussion was brought here primarily because this policy is where the 'source type' distinction was referenced from Notability. Since this policy, which should be strictly about No Original Research took it upon itself to create terms and definitions not directly related to No Original Research, another policy was linked here for these definitions/terms, hence this policy's discussions were yet again sidetracked. wbfergus Talk 14:37, 16 December 2007 (UTC)
Yes, this is why appending PSTS as a rider to this policy is a bad idea. It really has nothing to do with original research, because if there is a published primary or secondary source for material, it is not original. Dhaluza (talk) 17:40, 16 December 2007 (UTC)

Is research ever really defined?

It doesn't seem to be. This may be an important point because research means different things for people of differing backgrounds. In the sciences it's labwork and playing with data but in the humanities research often consists of reading secondary sources and abstracting them, something which is not unlike what's done in preparing a wikipedia article. It seems like in some cases this results in people talking past each other because they do not have a common conception of what they mean by research. Obviously, it's the bit about new synthesis of information that's important here but this is confusing because research does not typically involve synthesizing information at all. The synthesis is done when you're writing up the results of the research. The process of researching a subject for wikipedia and researching a subject for academic writing is pretty much the same. The difference is in the writing.

Isn't all research original btw? Even if you're reading someone else's notes, you're still doing your own research when you read them. --69.17.124.2 (talk) 18:50, 15 December 2007 (UTC)

Hi new user (judging by your sole use of an IP address). It's really a long convoluted process here on Wikipedia, sort of like trying to track just one string through the largest ball of string in the world, without unravelling the ball. Basically this policy states (or should state) that any thoughts, interpretations, synthesis, etc. must be published in a verifiable and reliable source first. Otherwise, if you can't cite a source that has those conclusions or other statements, it would give every appearance of being original thought (or research) by you, the editor. wbfergus Talk 19:06, 15 December 2007 (UTC)
The "research" portion of "No original research" (NOR) is a defined term, since WP can't actually prevent you from doing your own research. We simply require that you don't source the research you contribute here to yourself, but instead provide outside published reliable sources. That is where NOR should stop, but there is an activist position that NOR goes beyond using unpublished original sourcing, and should actually cover the use of material from published sources. That is one of the root issues of the current dispute here, and would make the "original" portion of NOR also a defined term. Dhaluza (talk) 01:48, 16 December 2007 (UTC)
Dhaluza, the policy's name "original research" is a bit misleading because it covers original thought, synthesis & argument as much as it does research. What this policy prohibits is a) novel interpretation of sources (including synthesis and re-readings) and b) new or "original" research which has not been published a reliable third party source. The former is what I understand you're calling the "activist position." As far as I'm concerned the prohibition on novel synthesis is absolutely necessary, new/personal interpretations of sources are by definition not the mainstream understanding of them, and recording mainstream understanding is what an encyclopedia does.
To answer the IP user no not all research is original research - neutrally summarizing the findings recorded in reliable sources is documentation of somebody else's research hence not original. Other than that I think wbfergus has addressed your points--Cailil talk 02:24, 16 December 2007 (UTC)
Nah, the position that Dhaluza referred to is going beyond the a/b you correctly described.
For that position, original research is not just what has not been previously published. For them, original research is governed by some special ingredient only known to the exalted. This special and magical ingredient, so the rationale of proponents of the PSTS section, gives primary sources (and only primary sources) a particular feature that sources in general don't have. Wait! Not only just primary sources, no, primary sources as covered by an arbitrarily chosen definition thereof.
Whatever this special ingredient is, it somehow makes it "necessary" to treat "primary" sources as distinct from sources. Consequently, its not just sufficient to say that no interpretations can be made from sources. Additionally we must also have a long winded discussion of primary sources too, and while we're at it, toss in a discussion of secondary and tertiary sources for good measure.
Perhaps all very "useful" to the aforementioned exalted. But normal drones, accustomed as we are to the strop of Occam's razor, we who have not been blessed with the royal jelly of queen-bee revelation, we can't understand it.
So, as you 'c', its not just the "stick to the sources" a/b definitions you and I and the whole wide world of non-'l33t drones knows about. -- Fullstop (talk) 04:15, 16 December 2007 (UTC)
Yes, the "research" part is a defined term that is used collectively to cover thought, synthesis & argument as you say. And I agree that new interpretations are not permitted, but only because they are previously unpublished and therefore original to the editor as far as WP is concerned. So the original part of original research should retain its plain language meaning. The activist position I am referring to is defining the original part to cover things that have been previously published, but in one type of source or another. Dhaluza (talk) 17:47, 16 December 2007 (UTC)

Secondary sources: bias, fairness, and objectivity

Vassyana recently made an excellent addition to the secondary source section:[22]

Secondary sources are usually preferable because they generally provide analysis and a broader context for the subject, and usually offer a more independent view. However, they can be biased by the views of their author(s) and/or publisher, so caution is required to preserve the neutral point of view. Care should also be taken to avoid undue weight and ensure the information cited is used in context.

SlimVirgin removed Vassyana's addition without any rationale.[23] In response to this removal of important, relevant (and easy to source) material, I added the following:[24]

Secondary sources on Wikipedia should have a reputation for fair and objective reporting, as they can be biased by the views of their author(s) and/or publisher.[4] Caution should be taken to preserve the neutral point of view , avoid undue weight and ensure the information cited is used in context.

SlimVirgin responded to my addition/modification of Vassyana's original edit by reverting to her version which removes the warning about secondary source bias. Her reason in the edit summary stated: "rv that isn't true -- secondary sources do not need a reputation for unbiased reporting; we have to use biased secondary sources all the time per NPOV)"

Read that again, please: we have to use biased secondary sources all the time per NPOV. This is a very strange assertion. There is nothing in the NPOV policy that is at odds with using fair and objective secondary sources. The nutshell states that "All Wikipedia articles and other encyclopedic content must be written from a neutral point of view, representing significant views fairly, proportionately and without bias." This is very clear. One is not required to use "unfair and subjective" secondary sources to represent a view -- it never needs to be done at any time. If a particular POV cannot be found in a reliable, secondary source that is considered fair and objective, then it probably does not meet notability guidelines, nor is it a "significant view" per NPOV. However, there are always exceptions: historical articles, articles about an organization, or important niche topics (such as scientific subdisciplines) may make use of primary sources carefully to represent a POV where there is a dearth of secondary sources, but in such cases, we should always strive to find fair and objective secondary sources whenever possible and question why they are lacking.

Wikipedia needs to maintain the highest ethical standards in sourcing while at the same time paying attention to NPOV. I fail to see how SlimVirgin's reversion and subsequent rationale is supported, and I find the removal of secondary source bias issues from this policy and the related ethical principles very troubling. Wikipedians must be ethical researchers, and we must start by acknowledging the deficiencies of secondary sources. Editors should be encouraged to evaluate all sources for fairness and objectivity, but on Wikipedia the concern is secondary sources; author, publisher, and editorial bias in primary and tertiary sources on Wikipedia is almost unknown. —Viriditas | Talk 08:52, 16 December 2007 (UTC)

While I agree that especially biased sources are problematic, I believe the change as worded will lead to wide swaths of perfectly legitimate sources being excluded. For example, it most cases one can hardly say that a Jesuit scholar writing about Catholic theology is "unbiased" or "objective", but certainly in most cases that would be a perfectly reliable and desirable secondary reference to use in describing Catholic theology. Does that concern make sense? Vassyana (talk) 12:34, 16 December 2007 (UTC)
Moreover, there is a vocal group of people who consider all mainstream science 'biased' because they view the scientific establishment negatively. So it's no use to try to exclude biased sources, since then this minority will try to use that language to exclude valid mainstream sources. — Carl (CBM · talk) 13:37, 16 December 2007 (UTC)
I can clearly see both side arguments in this instance. But, the actions without honest and active discussion by the parties involved speak more volumes. If you are going to add, edit, or revert, please have an active and honest discussion first. Otherwise your actions appear as knee-jerk reversion and smacks of ownership. This is a policy for all of Wikipedia, not just a select few to interpret and implement as they choose. wbfergus Talk 13:40, 16 December 2007 (UTC)

Vassyana, thanks for your reply. We agree about mentioning bias in secondary sources. Previously you wrote that SS "...can be biased by the views of their author(s) and/or publisher, so caution is required to preserve the neutral point of view." When this was removed, I added that SS "... should have a reputation for fair and objective reporting, as they can be biased by the views of their author(s) and/or publisher."

I'm not sure if you (or anyone else) saw what I did, but I encapsulated source evaluation and WP:RS into NOR. You see, Vassyana, not only are these three interconnected, but one should be able to see how every policy and guideline works together without having to read a separate page on each one. Think of it as summary style. That is to say, when the average Wikipedian visits a policy page, they should be able to come away with how that policy works - not only in and of itself - but in relation to each corresponding policy and guideline, all without having to visit multiple pages. This does not mean that legitimate sources will be excluded; it means that editors should be reminded to evaluate sources, and should also be aware of reliability issues. And, they get three for the price of one.

Now, as for your hypothetical example, you may find this hard to believe, but I've actually had to deal with a very similar situation. A group of editors (including myself) were writing about a science fiction film that made subtle allusions to Christian themes. A dispute arose between two editors over whether it was acceptable to use a religious source to criticize the film and at the same time discuss the religious themes. Although I wasn't a party to that dispute, the sources were removed. I remember glancing at them and seeing right away that they didn't pass a quick evaluation: the source did not appear to have a reputation for fair and objective reporting. As luck would have it, I was able to fill a void in the article by finding a similar piece of criticism by a Jesuit film critic who had a history of objective film criticism, and whose work was published by an academic religious journal with a reputation for fairness. So, it is possible to find reliable, fair, and objective secondary sources, but it requires the researcher to be vigilant, which is exactly the type of editor that we want. I believe I have met your concerns somewhat, although that will of course be for you to judge. But please, consider what others are saying. Turabian's 7th edition of A Manual for Writers (2007) says the following:

Beware, however, of online encyclopedias, such as Wikipedia, that rely on anonymous contributions rather than on carefully edited entries written by established researchers. Wikipedia has proved to be relatively accurate in the sciences, but overall it is uneven and sometimes wrong. Never cite it as an authoritative source.

We want to alert our contributors, registered or anonymous, about the hazards of secondary sources, particularly because Wikipedia relies on them more in non-science topics than on primary sources used in science. —Viriditas | Talk 15:00, 17 December 2007 (UTC)

I can't think of a situation where I would want to use an unfair or biased, source, but I can certainly think of a case where I would like to use one that's wrong. People keep sticking the statement that "AD does not mean After Death" into the Anno Domini article. I don't know if this is just vandalism, or if there are lots of people out there that think AD acutally stands for After Death. But in the absence of a good source that either wrongly claims AD acutally stands for After Death, or a good source that says this is a widespread misconception, I don't think we can say anything about it in the article.

Now, if I want a secondary source that's wrong, it isn't to hard to imagine that some other editor wants a secondary source that's biased, just to show that people exist who hold the biased position. --Gerry Ashton (talk) 15:27, 17 December 2007 (UTC)

But I can use a primary source to do just that, and in many cases, that primary source is supported by a fair, objective secondary source saying the same thing. There really isn't a good reason to use biased secondary sources at any time, although I've been told that "people do it all the time". That must make it ok, then...for eight million readers of Turabian to be told to stay as far away from Wikipedia as possible. If people want to keep supporting policies that are driving this site into the ground, that's their choice. —Viriditas | Talk 15:42, 17 December 2007 (UTC)
You can only use sources you can find. Your position is that the existence of a biased point of view must not be mentioned in an article, unless a fair and unbiased source can be found that says the point of view exists, no matter how many biased secondary sources exist that put forward the point of view. --Gerry Ashton (talk) 15:48, 17 December 2007 (UTC)
I never said a biased point of view must not be mentioned. I said: "Secondary sources on Wikipedia should have a reputation for fair and objective reporting". Huge difference. And biased POVs are often covered by reliable secondary sources that are considered fair and objective. What you're really saying is, what happens when you can't find a source that meets this requirement? This is already covered by WP:RS and the procedure for evaluating sources. I'm just putting it into words. —Viriditas | Talk 15:58, 17 December 2007 (UTC)

Editing comment

Somehow an inadvertant edit was made which wound up duplicating most of the content on this page, except for (I think) the last 5 edits by Smalljim and AnteaterZot. So try and fix this, I removed the first (original) set of the discussions, as I think the subsequent edits were added to the end, which would be in the duplicated section. I will notify those two users what I did and why, asking them check if I did this correctly, and to readd their comments if not. Any other problems anybody notices, please feel free to correct. wbfergus Talk 12:03, 16 December 2007 (UTC)

Thanks, I'd just noticed and was about to delete, but you pipped me to it!  —SMALLJIM  12:07, 16 December 2007 (UTC)
(Using my deepest Arnie voice) No problemo. wbfergus Talk 12:11, 16 December 2007 (UTC)

Recent changes to secondary sources

A lot of undiscussed changes were implemented into the recent replacement of the secondary sources paragraph, along with some edit-warring back and forth over it. A user has a conduct RfC active, an ArbCom case was filed and this page has been protected for weeks at a time. These symptoms arose exactly because of undiscussed changes and reverts. At this point, it's clearly and plainly disruptive behavior. Please discuss any significant changes before making them. The discussion for that text is at the obviously named WT:NOR#Secondary sources. Reviewing some of the changes:

  • [25] Ignores the reasoning behind the changes clearly linked in the discussion. Ignores the footnote which states that secondary references are "based upon primary sources and the work of other authors" [emphasis added].
  • [26] A significant alteration of the text, completely without discussion.
  • [27] Within the spirit of the proposal, though the introductory phrase is a significant clarification/specific point and should be discussed first. Posted for discussion by the editor at WT:NOR#Secondary sources: bias, fairness, and objectivity.

If someone wants to comment on an active discussion, they are perfectly welcome to do so. People waiting until after a change is implemented (based on talk page discussion) to impose their "discussion" through editing is simply intolerable and disruptive. If a change occurs after discussion and you disagree, please voice your opinions on the talk page like everyone else. Vassyana (talk) 12:29, 16 December 2007 (UTC)

Incidentally this episode is a point in favour of my suggestion of an "under discussion" tag for policy pages.[28]  —SMALLJIM  12:59, 16 December 2007 (UTC)
Outside of a single comment (which was addressed), those significantly altering and/or reverting the secondary sources section have not joined the discussion. As such, I am restoring the proposed draft since it lacks significant and/or substantiated opposition. Vassyana (talk) 17:04, 17 December 2007 (UTC)

Once again reverted by someone who has refused to discuss the changes, replacing them with undiscussed changes of their own. It's obvious a small number of people will express ownership over this policy. I could call exhortations to discuss already-discussed changes from those making significant undiscussed changes of their own a lot of less pleasant things as well. Vassyana (talk) 17:50, 17 December 2007 (UTC)

A reminder to all

As we continue to debate the wording of the PSTS section, there is one thing that I think needs to be remembered: The purpose of the secton is not intended to say that primary sources can never be used in Wikipedia! Even the current language makes it clear that primary sources can be used.... However, they need to be used with great caution, because it is very easy to misuse them in ways that introduce ORIGINAL RESEARCH into our articles. This to me is the key concept of the section, and is much more important than the exact definition of "primary" and "secondary". Remember, the whole point of the section is to prevent ORIGINAL RESEARCH... it is not to prevent reference to primary sources. The question we should be focusing on is not whether we define primary and secondary accurately... it should be whether the current language is an effeciant statement of this caution... in other words, does the current language properly caution editors about misusing primary sources in a way that introduces ORIGINAL RESEARCH? If so, then the definitions become a secondary issue (it does not matter if the source is primary or secondary if you have not introduced OR in referring to the source) ... if the current language is not an effective caution, then we need to re-write the section so it is. If we remember that this is all about NOR, and isn't really about the type of source, we should end up on track. It's that simple. Blueboar (talk) 14:17, 16 December 2007 (UTC)

The fact is that any source is subject to misuse. Whether something is "primary" or "secondary" or "tertiary" doesn't really matter to its potential for misuse. Consider a piece of journalism, an op-ed/opinion letter, a peer-reviewed scientific journal, a scientific review article, a textbook, a video taken by a citizen of an incident of police brutality and widely disseminated on the Internet. Each one of these is a source that can be misused by a wikipedia editor -- quoted out of context or misinterpreted -- and any misuse might be considered introducing original research. The statement about misuse ought not to be limited to primary sources. --Lquilter (talk) 14:46, 16 December 2007 (UTC)
(This statement about misuse in the guideline itself appears to be intended to provide an explanation within the guideline help explain of what the "problem" with primary sources is. But it was conceived with a particular idea of primary sources -- perhaps documentary videos -- and then expanded and applied outward. But identifying the "problem" with primary sources has run aground because this group cannot even agree on what the term means, and the more it talks about the issue the more confused everyone is. That suggests to me that PS/SS/TS is not an entirely helpful distinction. --Lquilter (talk) 14:55, 16 December 2007 (UTC)
I am not really trying to explain the problem... my statement is simply to remind us that any PSTS section needs to be geared towards preventing OR, not at preventing citation to primary sources. I think that, in all our discussion about definitions, we have lost sight of the fact that this is a caution and not a ban of Primary sources. If editors have paid attention to the caution, and not introduced OR into an article... then it does not really matter what type of source they are using. At least not as far as this policy is concerned.
Now... I think that the consesus is that we do need to caution people about misusing primary sources... and that we need to explain what we mean in that caution. Thus, some form of PSTS section is needed. However, I think the current language is so overly detailed that we lose sight of why it is there in the first place... to explain the caution. I simply wanted to remind people to keep that caution in the fore-front of their minds as they bounce ideas on language back and forth. Blueboar (talk) 16:34, 16 December 2007 (UTC)
Clarification: I didn't mean that you were trying to explain; I meant that the role of that point, in the guideline, was an explanatory one. Will clarify my own comment above. --Lquilter (talk) 18:26, 16 December 2007 (UTC)
Agree with your points. Kind of along these lines, it should be noted that the policy has had numerous edits the last day or two, going back and forth. The 'edit summaries' also indicate a bit of friction, with one comment saying the discussion was here. That user, Brimba, has made 6 edits to policy out of the last 50 edits. They have also had a total of 3 signed comments on this talk page. That's hardly a discussion by any means, unless everybody agreed with their sentiments, which the comments seem to indicate otherwise. Just an observation about the childish nature of some things going on here, and primarily by Admins, since we 'regular' users can't participate due to page protection (the lock icon is still present). wbfergus Talk 17:04, 16 December 2007 (UTC)
I do not currently have the sysop bit. I believe that the icon is there in error, or that it's semi-protection, as I can edit the policy. Vassyana (talk) 17:14, 16 December 2007 (UTC)
Actually it's move protection. (It uses {{pp-move}}.) Vassyana (talk) 17:18, 16 December 2007 (UTC)
Ah. Thanks for the clarification Vassyana. I just saw the lock icon and assumed it was still protected against us 'regular' users. I never knew moving (renaming) this policy was an issue. I thought it was (almost) entirely centered around PSTS. wbfergus Talk 17:23, 16 December 2007 (UTC)
Agreed with Blueboar's basic thrust starting in the opening comment of this talk section. It can be easy to lose perspective while seeking a suitable definition. I trust that with more participants persently involved, many of whom appear to me to have gained an increased understanding of the issues involved in PSTS, the result of the present round of debates will be reasonable and I hope fairly stable. At least, I hope so. FWIW, I just made a minor qualification to the presently used definition here. ... Kenosis (talk) 17:05, 16 December 2007 (UTC)
I don't see how the presence of those editing the policy without discussion and contrary to discussion is going to make anything "reasonable" or "fairly stable". I believe the result is quite the opposite. Vassyana (talk) 17:14, 16 December 2007 (UTC)
I think part of the problem is that we don't agree on the definition of "original research". The definition has been expanded over time to include all sorts of things that deal with the use of published sources, which has nothing to do with original material sourced to editors. Using citations for material that does not actually come from those sources is an original research problem. But if the citations are accurate, but the use of the sources is questionable, that is a reliable source issue, not an original research issue. I think that until we reach agreement on the scope of NOR, there is no way to agree on the definitions of PSTS. Dhaluza (talk) 17:24, 16 December 2007 (UTC)
So we are basically stuck in a stalemate since one 'side' wants the policy to strictly be about NOR and the other 'side' insists defining PSTS within a policy about NOR is an inherent part of NOR, and neither side seems willing to budge an iota on the issue, no matter how many 'discussions' take place. So, the only two alternatives seem to be to completely ignore it and invoke IAR when pressed, or to work on a way to make PSTS clearly understandable to almost every Wikipedian who comes to the policy for clarification. wbfergus Talk 17:34, 16 December 2007 (UTC)
I think the objections to moving PSTS to a separate guideline are based on keeping it attached to a policy, so it has greater weight in discussions. Moving it would allow us to focus on improving the NOR policy to directly address OR problems, without having to take on RS and V issues, which only makes the whole thing too convoluted to deal with effectively. That is the basis of the stalemate--a basic disagreement on the scope of NOR and its overlap with other policies. We have seen too many discussions about what is the legitimate mission objective of NOR diverted to say that PSTS is useful in content disputes and questioning it prevents good editors from confronting bad editors with it. So we can't discuss the basic policy elements without questioning the entire policy under this M.O. Dhaluza (talk) 18:07, 16 December 2007 (UTC)
For my part, if that is the main sticking point ("keeping it attached to a policy, so it has greater weight in discussions"), I wouldn't have any problem with a new policy that was strictly about PSTS. Wouldn't that accomplish the same thing and still address any of the issues that've brought up? The other policies could simply be adapted to say there was 'interaction' with the new policy as well. I have no idea how difficult is is to create a new policy though, maybe that's a stumbling block in itself. wbfergus Talk 18:13, 16 December 2007 (UTC)
If this is a proposal to move the PS/SS/TS material to a separate page, I wholeheartedly agree. That material can then be discussed and defined as needed. My own take is that PS/SS/TS is a helpful model in some situations, but not all. Separating these would immensely help, I believe. (And if we separate and a year or two from now it appears that they are creeping toward one another we can always re-merge.) --Lquilter (talk) 18:30, 16 December 2007 (UTC)
It wasn't offered as a proposal, but merely as a statement. We've addressed this several times in the past, but I 'think' maybe the wording then was just moving it into a guideline. If so, maybe this approach would be acceptable. wbfergus Talk 18:45, 16 December 2007 (UTC)
Directly adressing Blueboar's last comment, and reflecting subsequent comments by wbfergus/lquilter/dhaluza...
It is not the "PSTS section needs to be geared towards preventing OR." It is WP:NOR that is supposed to be geared to doing so. The PSTS section doesn't do anything to assist WP:NOR in achieving that end.
  • To quote the introduction to PSTS: "For the purposes of Wikipedia content policies and guidelines, primary, secondary and tertiary sources are broadly defined as follows." This is not a section geared towards preventing OR. This is a section in NOR, but about quite something else.
  • To quote the conclusion to PSTS: "All sources should be used in a way that does not give rise to (etc etc) original research."
  • Besides a token allusion to OR in the tertiary bit (ala "tertiary sources can ... be useful in preventing original research" or something equally non-brilliant), not once until that solitary sentence in the conclusion does PSTS even refer to "original research."
  • Whatever it is people think PSTS is doing, a relationship to NOR ain't there. It doesn't even have a reason to be there, because misuse of sources - irrespective of kind - is what NOR is supposed to be prohibiting. Thats what the conclusion also says.
Oh, I've checked SlimVirgin's edits for early October 2006, and understand full well what the psts stuff hoped to accomplish, and I think I can even recognize the thought processes involved (all good) that led to the instruction creep.
But irrespective of how PSTS "happened," the net effect was a) it was inserted into the wrong policy, b) it happened the wrong "way," and c) because it was in the wrong policy it got mangled to the cruft we have now.
A PSTS distinction is fundamentally unable "to be geared towards preventing OR." It doesn't do so now, and there is simply no way to make it do so in the future either.
-- Fullstop (talk) 19:01, 16 December 2007 (UTC)
  • I agree wholehearedly with Fullstop above. In response to the idea of moving PS/SS/TS to a separate "policy", I can't imagine how PS/SS/TS can achieve consensus as a policy when editors can't even agree on what the terms mean. I think it should be a guideline, and undergo a separate evaluation and discussion that assesses when it's helpful. But separate it to keep it from mucking up the WP:NOR material. To be honest, the PS/SS/TS reads as a guideline -- it's so detailed & specific. The core policies should be simple and clear to understand, and the murkiness of these distinctions & their applications really destroys that. Considering the policies to be a sort of Constitution for Wikipedia, they ought not be prolix but understandable. Anyone's MMV as to what constitutes understandable but certainly it is apparent by now that the tortured PS/SS/TS distinction is not understandable, either in text or in application. --Lquilter (talk) 19:10, 16 December 2007 (UTC)
A few thoughts about definitions. Because of the wide variety of Wikipedia editors, and the fact that many of them don't read, or only skim, policies, Wikipedia is not a good environment to try to create definitions that differ from the ordinary meaning of words. Also, once an organization (e.g. Wikipedia) creates a definition for one purpose, it tends to reuse the definition for other purposes. So saying that the definitions of primary, secondary, and tertiary sources is just for the purpose of preventing original research probably won't work. --Gerry Ashton (talk) 20:28, 16 December 2007 (UTC)
Agree wholeheartedly with your first points. I'm not sure what you mean by the last sentence. Do you mean to say that the defining PS/SS/TS in the NOR guideline won't prevent OR, because the definitions of PS/SS/TS are vague? --Lquilter (talk) 21:07, 16 December 2007 (UTC)
Let me revise my last point. The NOR policy currently states "for the purposes of Wikipedia content policies and guidelines, primary, secondary and tertiary sources are broadly defined as follows:". So the definitions of source types in this policy apply to every other policy and guideline too, and all policies and guidelines must be written accordingly. Alternatively, the beginning of the PSTS section contains a serious falsehood, which invalidates the entire section. --Gerry Ashton (talk) 21:22, 16 December 2007 (UTC)
I agree. If references to primary/secondary sources are to be made, as references to facts, then the references should be to the mainspace articles primary source and secondary source where they are expected to be NPOV, verifiable, etc, and any editor is welcomed in improving them. --SmokeyJoe (talk) 22:00, 16 December 2007 (UTC)
I would agree with moving the section on PS/SS/TS definitions to a guideline and relying more on examples that we can agree on, rather than definitions that we can't consistently apply. Tim Vickers (talk) 21:13, 16 December 2007 (UTC)
Wow... several people have taken my comment and run with it alot further than I intended. Just to make my position clear... I did not intend my comments to start another round of "let's get rid of PSTS or at least move it". We have been down that road, and a very vocal group of editors have made it clear that they think some form of PSTS is vital to this policy. That is not going to change. Everyone needs to accept this if we are ever going to get consensus language we can all live with. So, given that we are going to have some type of PSTS section... all I really wanted to to was remind people of the point to all of this. Any language we come up with should focus more on avoiding OR and less on avoiding Primary sources. That's all. Thanks Blueboar (talk) 22:36, 16 December 2007 (UTC)

Moved from bot discussion page

I'm afraid that Tim Vickers asked his question in the wrong spot, but as you can see, nobody there has come forth with a single reputable source that defines primary and secondary any different that I did. I'll paste my sources, which completely agree with me, below, with some annotation. Many of these are university library's pages. Do you think it likely that university librarians don't know what they are talking about? Or that science professors would allow the librarians to get it wrong so often, since they have to tell their students what a primary and secondary source is while directing their student's writing?

  1. University of Hawaii "How does Biological literature evolve over time?"
  2. William Madison Randall Library "Identifying Primary, Secondary, and Tertiary Sources"
  3. Cal State Northridge "Types of Resources"
  4. University of Minnesota Bio-Medical Library "Primary, Secondary and Tertiary Sources in the Health Sciences"
  5. Diablo Valley College "Identifying Primary, Secondary and Tertiary Sources"
  6. Florida Institute of Technology "Primary, Secondary, and Tertiary Sources"
  7. Yale University "Lesson 3: Recognizing Primary, Secondary, and Tertiary Sources"
  8. OhioLINK, is a consortium of 86 Ohio college and university libraries, and the State Library of Ohio "Cycle of Scientific Literature"
  9. University of California Santa Barbara "Distinguishing Between Primary, Secondary and Tertiary Sources"
  10. eric.ed.gov Education Resources Information Center "Chemical Literature Exercises and Resources"
  11. And finally, a book: Writing Guidelines for Postgraduate Science Students By John Hampton. AnteaterZot (talk) 10:19, 16 December 2007 (UTC)

I moved this from User talk:ProteinBoxBot, since that isn't really the place to discuss the details of this policy. However, as I read the discussion above, the problem seems to be that some people are arguing that the Wikipedia PS/SS definitions do not follow those used in other areas. Speaking personally, I am very unhappy that we have a policy that is not interpretable since there is no consensus on the meaning of the words used. This is becoming increasingly disruptive to the project. Tim Vickers (talk) 20:42, 16 December 2007 (UTC)

  • Completely agree. As an editor who simply came to these pages originally to seek some guidance for how to explain a problem to users on a page, I found the pages to be more confusing than helpful. Then when I started reading the talk pages to see if that would help it became even clearer to me that not only is the PS/SS/TS not helpful on the policy, it is an incredible source of disruption and argument. --Lquilter (talk) 21:13, 16 December 2007 (UTC)

Moving PSTS

Any attempt to move PSTS out of this policy will be soundly rejected by the same group of editors who seem to think that having all of these circular discussions, resulting confusion and edit-wars about source-types is beneficial to both this policy and Wikipedia. It's easy enough to rattle off the names by now just from memory, but it almost as easy for any reader to breifly peruse this talk page, the last couple of archives, and view the edit history of the policy itself to see who would oppose it. This also at least would give the casual reader a context for some of their positions, which I wouldn't be able to accurately define just from memory. Some of the comments by those in the group that oppose moving PSTS out of this policy include statements such as:

  1. ...an attempt to weaken policy...
  2. ...an attempt to derail policy..
  3. ...the people who are most frustrated by this policy are those who want to violate it. Well - good!!
  4. a non-starter
  5. ...but major changes such as removal of material that has been stable for years, cannot be contemplated unless there is wide support.

etc.

So, as I stated in a section above, what we are left with is a stalemate of being left more-or-less with something that is confusing to most Wikipedians. This policy talk page will continue to be almost exclusively about PSTS, and edit wars will continue on the policy itself into perpetuity. This then begs the question. Since we are stuck with these ill-defined terms that are used with different meanings in different disciplines in different contexts, how can (or can it) be corrected within this policy?

It should be noted that back around August or so, Vassyana was one of those who opposed many of the changes being discussed. As any casual reader of the talk page in the last few weeks or longer can tell, that position was changed and Vassyana has worked extremely diligently towards compromise language acceptable to all parties, but we still have the minimal definitions which has helped, but not eliminated the 'confusion factor'.

Back in September, I stated "Also, I would tend to not go with the self-contained approach, as I see that as what's caused this mess. The policy itself does state "Since the policies complement each other, they should not be interpreted in isolation from one another, and editors should try to familiarize themselves with all three.", it seems to me as if the self-contained approach has already been addressed as not viable." This is in regards to each policy having its own self-contained 'blurb' about sources. This was followed by a comment from one of those currently in the group opposed to (most) changes with "I agree with you on the 'self-contained approach' wbfergus. Due to the nature of these sourcing and content policies, there has to be some overlapping areas between them, we just need a better means of making sure that the overlaps match, and are not confusing or contradictory to each other". So, there is some minimal ground available for a compromise amongst all parties, but how do we reach it?

I could see the entire PSTS section being greatly expanded to cover any and all defintions as used by the various disciplines, which may reduce the sheer amount of volume of discussions on this page, but at the same time it would make the policy itself extremely unwieldly and far more difficult to read and understand. Most readers would quickly be consumed with "What does this have to do with original research?" or after spending a huge amount of time reading and digesting it, would say "Okay, now I understand what original research means, and I also know how every other discipline uses the terms primery, secondary and/or tertiary sources".

I could also see the PSTS section remaining pretty much the same, with only minor changes over the years, as one 'group' thinks the policy is fine as it is, they don't bother much with the discussion page anymore, and will constantly revert any changes to the main policy page no matter how many people agree to a change on this page.

Does anybody else see any other alternatives? wbfergus Talk 23:03, 16 December 2007 (UTC)

I would certainly support moving PSTS to its own self-contained guideline. It would allow the kind of space appropriate to examine the different general subjects and situations, as well as address the issues of neutral point of view, what Wikipedia is not, verifiability, reliable sourcing and notability that are related to the source distinctions. It would also do a lot to stabilize this policy. An alternative would be work on an alternative but parallel paradigm (essentially a terminology change). A lot of baggage comes with the primary/secondary terms. I'd personally recommend a distinction between what sources articles should principally rely upon (especially for analysis and determining NPOV) and generally "everything else". (You're welcome to some grains of salt.) Vassyana (talk) 23:14, 16 December 2007 (UTC)
I oppose moving it. It's central to this policy that articles rely on secondary sources, particularly for anything controversial, in BLPs, and to establish notability. SlimVirgin (talk)(contribs) 23:26, 16 December 2007 (UTC)
As I stated in my opening sentence/paragraph. So how can we address clearly up the confusion factor and endless discussions on this policy so that almost everything doesn't revolve around PSTS issues? wbfergus Talk 23:43, 16 December 2007 (UTC)
Slim, how exactly is it "central to this policy that articles rely on secondary sources"? I've seen this stated before but as yet no one has been able to explain it. Also, could you please explain this in light of the A reminder to all that was active until a few hours ago?
Thanks. -- Fullstop (talk) 23:51, 16 December 2007 (UTC)
It's been in the policy for about three years and is widely accepted. You need secondary sources to establish notability of the subject, and to establish that you're not engaging in original analysis or synthesis of primary sources. Articles with no secondary sources are frequently deleted. Can you show me a good article that contains no secondary-source material? SlimVirgin (talk)(contribs) 00:03, 17 December 2007 (UTC)
Having been in the policy "for three years" is not a reason that it is "central to this policy that articles rely on secondary sources." Also, the use of "secondary sources to establish notability" is not relevant to this policy. So, would you please answer the question? Thanks. -- Fullstop (talk) 00:08, 17 December 2007 (UTC)
Of course notability is relevant to this policy. You show that your ideas are notable, and not your own mish-mish of primary sources (or, worse, no sources), by producing reliable, published secondary sources that say what you are saying. This shows that other people think the issue is worth writing about, not just you. SlimVirgin (talk)(contribs) 01:16, 17 December 2007 (UTC)
Um, this is NOR policy we're talking about, right? Notability is about suitability of article/topic(/sources). "No original research" is about accurately reflecting a source statement. It is the antonym of "original research" which is novel interpretation (of a source).
Neither process influences or is influenced by the notability/reliability of the source that is/isn't being mangled. -- Fullstop (talk) 01:34, 17 December 2007 (UTC)
  • Re: SV's comment: But, since "secondary sources" is clearly confusing for people (since the usage varies across fields), why not just replace that with "reliable published sources". We could then link to the WP:RS page which discusses how to evaluate and assess sources, and maybe the material on source classification would find a better home here. (Wouldn't it be nice if some of this dispute could be resolved simply by recontextualizing the information?) --Lquilter (talk) 00:10, 17 December 2007 (UTC)
Because they must be reliable secondary sources, not just published. As I said, articles without them are routinely deleted. This policy has to be both prescriptive and descriptive of best practice. SlimVirgin (talk)(contribs) 00:24, 17 December 2007 (UTC)
Slim, I really do understand/appreciate the importance of SS for notability, but NOR is not prescriptive of notability. While a discussion of SS/PS may have merit in notability (though a link to secondary sources might also suffice), such a distinction of source type has nothing to do with the accurate portrayal of what sources say. --- 00:34, 17 December 2007 (UTC)
But NOR is (or purports to be) prescriptive of the meaning of primary source, secondary source, and tertiary source in WP:N and every other policy and guideline because it includes the phrase "for the purposes of Wikipedia content policies and guidelines, primary, secondary and tertiary sources are broadly defined as follows:". --Gerry Ashton (talk) 00:43, 17 December 2007 (UTC)
Zigzagtly! Nary a word about "original research," but plenty to say about notability, rs and other stuff. Useful? Perhaps to notability, rs and other stuff, but not to "original research." -- Fullstop (talk) 00:48, 17 December 2007 (UTC)
Well, they must be something other than merely published, but since there's no consensus as to what "secondary" means, then "secondary" maybe isn't the best way to describe what they must be. "Reliable" could summarize that, no? And refer people to WP:RS to ascertain what reliable means. PS/SS/TS then helps explain reliability. --Lquilter (talk) 00:28, 17 December 2007 (UTC)
WP:RS continues to be an bone of contention and that needs to be resolved before we can fix this. ≈ jossi ≈ (talk) 00:12, 17 December 2007 (UTC)
Is there anybody here that believe that an article can be solely based on primary sources? I so, please say so... ≈ jossi ≈ (talk) 00:13, 17 December 2007 (UTC)
Please don't change the subject. The subject "Moving PSTS" out of NOR. The subject is not RS or notability. Thanks. -- Fullstop (talk) 00:20, 17 December 2007 (UTC)
Do not tell me what to say or not to say, and spare us the "thanks". ≈ jossi ≈ (talk) 00:22, 17 December 2007 (UTC)
Jossi said WP:RS continues to be an bone of contention and that needs to be resolved before we can fix this. -- Jossi, what I'm wondering is whether moving PS/SS/TS to WP:RS would help both pages. First, it seems to be a better "fit"; second, the disputes here about PS/SS/TS have been over how to define and use the various categories; particularly, how does peer-reviewed literature fit here, when is it primary and when secondary. The disputes have not been over people putting in OR into wikipedia articles -- everybody agrees that that's not correct. All the squabbles are about what are the best sources. So since this is principally an argument about sources and use of sources, and that's what PS/SS/TS is most helpful for, it seems that moving it to WP:RS would clean up WP:NOR. And the PS/SS/TS material might offer another way to structure WP:RS to provide the best, clearest guidance for WP editors. --Lquilter (talk) 00:39, 17 December 2007 (UTC)

<<< Vassyana, would you undertake starting a sandbox for a PSTS page based on your comment above, as well as a summary that coukld be added to NOR? That way we well be able to evaluate the merit of that proposal. ≈ jossi ≈ (talk) 00:26, 17 December 2007 (UTC)

I believe I can do that. Give me a day or two to work out a first decent draft. Vassyana (talk) 17:25, 17 December 2007 (UTC)

Following a couple of comments (including mine) from which it isn't evident whether the commentator agrees or disagrees with a move, can we get this discussion back on track? Should PSTS be moved?
The lack of relevance to NOR is prima facie (even worse so since a few hours ago) evident. So far the only reason for an opposition to a move is because it is relevant to NOTE, which of course is then a reason for a move to NOTE. Any other reasons why PSTS should not be moved? -- Fullstop (talk) 14:55, 17 December 2007 (UTC)

Fullstop (and others). Personally I would full heartedly agree with moving PSTS somehwere more appropriate, but it simply ain't gonna happen. As I said in the opening paragraph, there are those who would oppose it, regardless of any arguments, evidence, etc. This 'block' consists of approximately 13 or more editors, so simple math (and a little 'guessitmation' on what makes a clear consensus) means for an 80% 'consensus', we would need around 50-60 other editors to agree to a move. Then the other group would bring in more editors, etc. It would be a very long and tedious process. So, since moving the section is nigh onto impossible, at least in the short term. what can be done to address the fundamental problem, which to me is primarily with the definitions within PSTS. I'm going to start a new section with an idea that may be agreeable to all parties, barring just obstructionism. wbfergus Talk 15:11, 17 December 2007 (UTC)
I'm curious as to why you say a move "ain't gonna happen" when there appears to be a rough consensus for a move. —Viriditas | Talk 15:13, 17 December 2007 (UTC)
As I said, for a roughly 80% consensus (and actual numbers have never been defined anywhere for what is a consensus), the 'block' supporting a move needs 4 people for each person in the 'block' that would oppose a move. After 4 months or so, there are only around 65 'participants' on this page, even after announcements on other policies, the Village Pump and a RfC. That in itself seems insufficient for something like an outright removal of material from a policy, unless perhaps something around 95% of all participants agreed. And again, that won't happen. I know it's awfully long and convoluted, but the best bet to see the 'reasoning' would be by reading through the Archives for the last 4 months, or perhaps a very general overview could be garnered by looking at the edit history of the policy, and reading the 'edit summaries'. There's also the ongoing RfC against COgden with many comments, and the ArbCom request with more additional comments. wbfergus Talk 15:33, 17 December 2007 (UTC)
Unfortunately, I've seen it all (and a lot more) and I'm less than impressed. There's a lot of history here you may not be familiar with. As for outright removal, the instruction creep and policy bloat has reached a point where if you removed half of it, it might be a great improvement. KISS. —Viriditas | Talk 16:12, 17 December 2007 (UTC)

Move PSTS to reduce churn

I'd say PSTS should be moved simply to reduce churn in the NOR talk page. This is really getting ridiculous. New section started after my fourth edit conflict trying to reply to the previous debate. Squidfryerchef (talk) 00:28, 17 December 2007 (UTC)

Go ahead and move it, if no prohibition in NOR depends on the definitions contained in the moved PSTS. Also, remove the claim that the definitions apply to all policies and guidelines, and restrict the scope to just NOR. If prohibitions anywhere in Wikipedia depend on definitions in PSTS, the churn will continue. --Gerry Ashton (talk) 00:33, 17 December 2007 (UTC)
No, don't. Start a PSTS page and propose it. ≈ jossi ≈ (talk) 00:35, 17 December 2007 (UTC)
I strongly think that all the established editors who have been mucking about with these pages should stop. That means editors from all sides. If people have to argue on these pages until the cows come home, fine, but the back-and-forth reverting on the main page is silly, and more appropriate to novice WP editors than to the established editors and admins that are doing it.
As for process going forward, since there are multiple ideas on the table, I suggest a page that lists the several major approaches -- it should be one page because the problem and solutions span multiple pages. (WP:NOR and WP:RS at the least; WP:FICT also springs to mind as an elephant in the room.) I set up an outline at User:Lquilter/Approaches and will undertake to fill in some pieces. --Lquilter (talk) 00:42, 17 December 2007 (UTC)
Hear hear, I second Lquilter's call for a "cease fire" here. I'd also echo what Jossi said - if anything is going to happen to PSTS it needs to proposed properly--Cailil talk 01:39, 17 December 2007 (UTC)
NOR, FUBAR. Tim Vickers (talk) 04:15, 17 December 2007 (UTC)

Why are people upset with this policy?

Let me suggest a reason people may be so upset about this policy. Original research is wonderful. A scientist who freezes her butt off to take climate data in Antarctica feels she is doing a service to humanity. Try to imagine how a scientist feels after spending years poring through data, and finally discovers a theory that puts it all in perspective. What would we know about the world if there were no journalists who go into dangerous areas and interview people who don't know if they are going to survive the day?

Granted, Wikipedia isn't a publisher of original research. But often in this policy, and on this talk page, original research is a label that is slapped on all kinds of sloppy or deceitful editorial practices, such as quoting out of context, or synthesis of ideas that misrepresent the sources. This is an insult to people who do real original research, and no wonder they get hot under the collar when something so important to them is equated to sloppiness and/or deceit.

To avoid insulting original researchers, I suggest that whenever possible, sloppy or deceitful practices be described in appropriate negative terms, such as undue weight, or quoting out of context, and not be labeled with the positive term original research.-Gerry Ashton (talk) 03:30, 17 December 2007 (UTC)

It is highly unlikely that any truly interesting result will not get commented on by any scientist (by means of a review article) and/or also not get picked up by any journalist (they have science writers on staff at many newspapers, magazines, and radio and TV shows. NPR has a Science Friday show.). Scientists can send out press releases, and do, all the time. AnteaterZot (talk) 03:50, 17 December 2007 (UTC)
The above point may be valid for scientific research where there is intense competition to publish the new. Given today's IP environment few self-respecting academics would want to put real research out into the public domain. However, in addition to the nut cases, religious zealots, political pushers, etc, Wikipedia does have some well trained editors who are perfectly capable of doing good original work. For these editors I think we do need to be careful about turning OR into a derogatory term.
These editors sometimes stray into the area of original research because it takes time to learn the difference between analyzing and assessing sources (as one might for a term paper, thesis, or advanced seminar) and merely reporting on them (as one must in a Wikipedia article or might in lecturing for an introductory survey course). These kind of editors don't always see their work as original research because the kinds of deductions they make would not be original research in a literature, philosophy, or religion department. To the contrary, to people in the field, they would simply be reasonable opinions consistent with the sources and not nearly "original" enough to merit publishing. They are only OR in the Wikipedia context where we want material to be verifiable by any reasonably intelligent high school student. Egfrank (talk) 13:46, 17 December 2007 (UTC)
GA's point To avoid insulting original researchers, I suggest that whenever possible, sloppy or deceitful practices be described in appropriate negative terms, such as undue weight, or quoting out of context, and not be labeled with the positive term original research. raises something else I've been thinking about, which is that we may be trying to make this one policy do too much. NOR and other policies are sources for the overall spirit of wikipedia, but they do not define everything -- that's why we need guidelines to provide more specific guidance for particular situations. People sometimes cite to NOR or NPOV to try to add more rhetorical weight to their argument, when in fact there is a more specific and appropriate point to make (like "undue weight" or "quoting out of context"). Sticking closely to the issue in discussions (here as on mainspace talk pages) could help keep the framework clear. Cheers, Lquilter (talk) 20:25, 17 December 2007 (UTC)

SlimVirgin's recent edit(s)

In an undiscussed set of edits, Slim Virgin removed the following passage from the policy page:

All sources should be used in a way that does not give rise to new analyses, syntheses or original conclusions that are not verifiable. Where interpretive claims, analysis, or synthetic claims are included in Wikipedia articles, use appropriate sources rather than original analysis by Wikipedia editors.

The removal occurred during an edit that was ostensibly because "someone removed that articles need secondary sources."

I question its removal,

  • because this paragraph is a key statements in NOR policy,
  • in light of SlimVirgin's assumption (?) that its necessary to ramble on about PS/SS.
  • that there was a discussion about this only yesterday, to which she did not contribute.

If nothing else, this is the application of a double-standard, i.e. "bold" editing, ignoring so-called "consensus", and the idea that one keeps thing because they've been in the policy since Noah.

Comments? -- Fullstop (talk) 14:46, 17 December 2007 (UTC)

Well, what I would like to see is the community enforce a 1RR on policy pages and ban tag-teaming reverts altogether. —Viriditas | Talk 15:22, 17 December 2007 (UTC)
Hear, hear. Or at least temporarily. (Although difficulty in defining "tag-teaming" - blech. Ideally editor/admins would just police themselves.) --Lquilter (talk) 15:47, 17 December 2007 (UTC)
I've undone the edits, because they were significant and it was not discussed on the talk page. What's good for the goose is good for the gander. Vassyana (talk) 16:58, 17 December 2007 (UTC)
I edited conflicted restoring Brimba's harmonization because SlimVirgin restored that portion of the revert (appropriately so). Vassyana (talk) 17:01, 17 December 2007 (UTC)

Would this be acceptable, re: PSTS

Perhaps a way towards some sort of compromise would be by leaving the 'Sources' section, but relegating the actual definition of the terms in the appropriate articles after those three articles Primary sources, etc. were expanded to have discipline specific definitions (maybe a lot of work?), each of them upgraded to at least 'guideline' status, and then appropriately linked from this policy (and any others), similar to "definition of these terms is beyond the scope of this policy, see the appropriate guideline for a more complete, discipline specific set of definitions". Most of the current wording would still remain in the policy, but the actual definitions are just moved elsewhere for more complete coverage. This policy could then merely state what it needs to about primary, secondary, etc. without getting bogged down in conflicts over what is a primary source, etc. in which context or in regards to which discipline.

In the interests of trying something new for clarity, can anybody responding please respond only in the correct designated 'area', either 'For' or 'Oppose' with a brief statement of why you think so? If you disagree with somebody elses statement, please do not respond directly underneath it, it makes following discussions for either 'side' more difficult. Use the separate 'Comments' section instead.

For

  1. wbfergus Talk 15:25, 17 December 2007 (UTC) agree (uh, because it's my suggestion?)
  2. Tentative support (subject to withdrawal at any time, of course. <g>) This would be a step in the right direction, because it would permit close examination of discipline-specific issues with respect to the definitions. The definitions should also include use scenarios which would actually make them much more helpful and understandable to the novice. Caveats: (1) I'm not sure that each of the source types needs a separate page--they might be better done all together as they are here. (2) "Promote" to guideline presumes that's the direction and that's good but it shouldn't be automatic. From what I can tell at this late date there is not (currently) consensus on these terms. (3) I believe that some of the difficulty is also in the prefatory material which describes how to use the source guidance. That will continue to be a bone of contention, I suspect. My own view is that PSTS is a helpful model but there will always be situations in which sources don't quite fit the model; it should be explained as a model so that people understand that we don't care about the classification, per se, we just care about the underlying principle -- NOR -- and he classification is one shorthand way to help figure out whether you have OR or not. --Lquilter (talk) 15:59, 17 December 2007 (UTC)
  3. Partial support. Although I don't think this is actually the right way forward, I would support moving the PSTS definitions to their own guideline page. Anything to get the logjam moving! As you should know, if you've read everything on this page :) I'm really in favour of encouraging a slow shift to using new terms.  —SMALLJIM  19:54, 17 December 2007 (UTC)

Oppose

  1. -Gerry Ashton (talk) 15:36, 17 December 2007 (UTC) Primary source is an article, not a guideline or policy. It should not contain any material that is only useful in making Wikipedia editorial decisions. If a parallel guideline were created, whether or not a given article or edit satisfied NOR and NOTE would depend on the interplay between the definitions and the policy. As a general rule, strict definitions require loose policy, to allow the use of sources that happen to fall into a definition that does not reflect the true character of the source. The slightest change to the definitions might require extensive rewrite of the policies, but when they are in separate pages, it is unlikely this would happen. --Gerry Ashton (talk) 15:36, 17 December 2007 (UTC)
  2. Oppose per Gerry. I'd much rather see something like Jossi proposes below. Dreadstar † 22:24, 17 December 2007 (UTC)

Comments

  1. wbfergus Talk 15:25, 17 December 2007 (UTC) This is totally non-binding. I'm merely offering this approach to see if there might be some common, middle-ground that most parties involved can agree might be a workable approach to all of the edit-warring and endless discussions. If there isn't any middle-ground on this suggestion, then we can always take a look at other ideas, but lets see if there are any areas of agreement in the 'small step' approach.
  2. If I understand correctly, this proposal is to put the source definitions into "guidelines" drawing from the articles to start with; is that right? If so, you might want to clarify the proposal to avoid the confusion that Gerry Ashton has had. I believe from his comment above that he thinks the proposal is simply to refer people to the mainspace articles, and that's not my understanding. --Lquilter (talk) 15:55, 17 December 2007 (UTC)
  3. I don't oppose the change, but I think it would be counterproductive, since what we are trying to get at is not primariness or secondariness. The central issue is interpretiveness. Quote mining is just as easy from secondary sources as from primary sources, however defined. But you can't quote mine from non-interpretive sources. Likewise, if you cite an interpretation of raw scientific data, it's a good thing, whether that interpretation is from from a primary or secondary source. COGDEN 19:49, 17 December 2007 (UTC)

Another reminder

It might be helpful to interject some history and remind people how the term "Primary Source" first entered this policy... It occured back in February of 2004, in this edit.

To me, the key point of the addition was to highlight that Wikipeda should not become a primary source... It did not talk about Wikipedia articles not using primary sources. In fact, for quite a while the policy specifically stated that you could use primary sources: as this version from early August of 2005 demonstrates: Original research refers to original research by editors of Wikipedia. It does not refer to original research that is published or available elsewhere (although such research may be excluded if editors consider the source to be disreputable or inappropriate).

Now, I do understand that policy changes over time... and what was included in 2005 is not going to be the same as it is in 2007. My point in showing these old versions is not to say we should turn back the clock to 2005... my point is to highlight the intent behind the introduction of source typing... the fore-runner to the current PSTS section.

Until quite recently, the entire point of discussing source types was focused not on the sources we were using, but on what kind of source Wikipedia is. Introducing OR into an article turns Wikipedia into a primary source... and Wikipedia should not be a primary source. This is a concept that I hope all of you would agree is at the core of the NOR policy.

The thing is... somewhere along the lines we lost the tie between source typing and what Wikipedia is and is not... we lost the statement that tied source typing directly to the concept of NOR. We lost the intent behind discussing source typing... the idea that Wikipedia is not a primary source. We now have a lot of explanation about source types, but no direct link to why they are important to the concept of No Original Research.

If we are going to keep the PSTS section... I think we need to add this back. We need to revive the statement that we do not want Wikipedia to be a primary source. Blueboar (talk) 15:55, 17 December 2007 (UTC)

  • This is a very good point. I think it helps deal with my concern which is that we do not care about classifying sources just because we care about classifying sources; but we care about classifying sources only because it is a means to an end -- a helpful shortcut to identifying whether or not something is OR (verboten) or sourced, reliable, verifiable information previously published in a reliable source (bring it on so we can neutral-ify it). --Lquilter (talk) 16:02, 17 December 2007 (UTC)
  • Where does it say in the current wording that Wikipedia articles should not use primary sources? It says: 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. which is very different. ≈ jossi ≈ (talk) 16:10, 17 December 2007 (UTC)
Jossi, I fully agree that the current wording does not bar primary sources... that it only cautions editors to take care in using them... I did not mean to imply otherwise (in fact, I pointed this out in my other reminder section, above).
I am addressing a different issue here: Why does this policy contain a section that defines Primary and Secondary sources in the first place? The answer is obvious when you look at the edit history: It was to explain what we mean when we stated "Wikipedia is not a primary source". That is a context that has gotten lost in the shuffle. It makes much more sense that we have to define what a primary and secondary source is when the section is placed in its original context. That single sentence ties PSTS to the rest of NOR, something that is somewhat lacking in the current language.
It also goes to intent... the concept of PSTS entered this policy to explain something about Wikipedia itself, not the sources we use in Wikipedia. We have lost that intent. We need to return it. Blueboar (talk) 17:00, 17 December 2007 (UTC)
Yeah, Jossi. Reiterating the "where does it say" inanity is now patently a case of "hello, are you not listening?" Well, even if you can't/don't want to hear, its been said, over and over and over and over again. So change the record.
  1. There is no need to say something that does not say something.
  2. yadda, yadda, "may" yadda "only with care" yadda. As if editors - in their own minds - ever supposed that they something without "care." From the point of view of the editor, he/she may use PS. Of course its with "care" of course he/she would never "misuse" them.
Oh, and please try to not indulge in non-sequiturs and other inconsequentialities. Although your intention is of course to distract/muddy the waters, it tends to also cast doubt on your ability to follow conversation. Thanks.
-- —Preceding unsigned comment added by Fullstop (talk • contribs) 17:16, 17 December 2007 (UTC)
It is below my dignity to respond to your provocation, and I will not. Have fun with your soliloquizing. ≈ jossi ≈ (talk) 17:21, 17 December 2007 (UTC)
  • Blueboar, yes absolutely the key point of the addition was to highlight that Wikipeda should not become a primary source. But this is the only relationship between OR and kind of source. For which it is not necessary to have an essay in policy about what primary sources are.
Further, the idea being expressed by that key point could itself be written in a way to avoid the dependency on potentially unfamiliar terminology. The idea being expressed is: don't introduce novel ideas that someone else might use for his/her research.
Thus, "the key point" could just as well be written as "A Wikipeda article may only contain material that has already been stated elsewhere."
When a key point is using jargon, replace the jargon with plain language, don't explain the jargon.
-- Fullstop (talk) 17:19, 17 December 2007 (UTC)
:When a key point is using jargon, replace the jargon with plain language, don't explain the jargon. (Fullstop) -- Yes. This is absolutely correct. --Lquilter (talk) 17:34, 17 December 2007 (UTC)

A line big enough to drive a truck through....

I just saw this on the article page:

Appropriate sourcing is a complicated issue, and these are general rules. The decision as to whether primary or secondary sources are more suitable on any given occasion is a matter of common sense and good editorial judgment, and should be discussed on individual article talk pages.

I understand the need not to pin people down, but the above line is tantamount to saying "Ignore all of the above". This seems to me more likely to aggravate disputes to the point of unresolvability than to clarify things. Egfrank (talk) 16:20, 17 December 2007 (UTC)

So be it. It is better to have aggravated disputes on talk pages than a guideline that inflexibly prescribes bad advice. And anyway - "ignore all of the above" is already policy. But most importantly, it's true - sourcing is complex. We teach entire college courses on appropriate selection and use of sources. We cannot hope to have a single policy page that gives advice that is universally usable. It is far preferable to encourage the use of discretion and common sense within a set of principles, and I am glad that we have finally adopted that approach. Phil Sandifer (talk) 17:17, 17 December 2007 (UTC)
I understand and agree sourcing is complex - people don't just take courses, they get entire PhD's to learn how to properly source and analyze work in a field. However, the above line negates the principles themselves and essentially turns common sense into a matter of personal opinion.
As I see it the purpose of these policy statements is to give the academically trained (the people who have taken those "entire college courses") at least a fighting chance of being heard in the midst of all the others - which includes very articulate kids and quacks along with the academically trained. Egfrank (talk) 00:22, 18 December 2007 (UTC)

Primary vs. secondary based on content

I'm wondering when a secondary source becomes a primary source based on what it contains (disregarding the identity or affiliation of the author).

  • Certainly secondary
    • Contains anything that would be proper in a Wikipedia article
    • Contains interviews or summaries of interviews with identified persons
    • Contains correspondence with identified persons
    • Author made decisions about what to include or exclude
    • Author decided which of several conflicting sources was correct
    • Contains excerpts or summaries of documents not available to the public
    • Contains calculations that could be reproduced by a person with specialized knowledge, starting from the same sources used by the author
    • For reviews, expresses the authors like or dislike of the work under review.
  • Possibly primary
    • Author proposes theory or explaination that (purportedly) harmonizes various sources, but is not contained in any of the sources
    • Contains statements about what is or is not the mainstream view, presumably based on extensive reading, but without giving a specific list of citations.
    • Expresses authors opinion about documents or objects, based on reading other sources (e.g. "there isn't a chance in hell this new standard will be widely adopted.")

—Preceding unsigned comment added by Gerry Ashton (talk • contribs) 17:32, 17 December 2007 (UTC)

That's just not how the terms are used in the sciences; the peer-reviewed articles in Nature, Science, etc., are "the primary literature". --Lquilter (talk) 17:37, 17 December 2007 (UTC)
I don't know about Nature and Science, but some science and technology journals publish some tutorial or review articles, which would be secondary sources by anybody's definition. --Gerry Ashton (talk) 17:44, 17 December 2007 (UTC)
Yes, but those aren't "the primary literature", nor are they the bulk of scientific publishing. Peer-reviewed articles publishing research are the "primary literature"; so-called "review articles" are the "secondary literature". Which is exactly why the PSTS is weird, to say the least, for application in science articles on wikipedia. ... Science is just one example of how the current definition is over-simplified and not adequately interdisciplinary. Another example is legal scholarship. Setting aside the question of legal documents (which typically include a significant amount of analysis themselves), legal scholarship has its own rules. The PSTS isn't inapplicable, per se, but other features peculiar to legal scholarship are arguably as or more significant. Peer-reviewed journals in law are few and far between; the vast majority of legal scholarship is published in student-edited publications -- and there are hierarchies within these, e.g., law review versus non law review journals. The critical distinction, for instance, between student "notes" and faculty publications in a law review is indicated within law simply by an italicized Note after the student name, and that significantly affects the reliability of the source. Etc. ... The current policy would suggest that citing a peer-reviewed paper in Science is less favorable for a proposition than citing to a review article published in, say, Tiny Journal of Miniscule Detail. So for DNA we wouldn't cite to Watson/Crick we would cite to Random Nobody. That's weird for a lot of reasons (not least because the review article is likely not peer-reviewed itself, and so its re-statements of the proposition have received less scrutiny than the original statements.) --Lquilter (talk) 18:49, 17 December 2007 (UTC)
I don't think there is discrepancy between what I wrote and what would be considered secondary in science. If someone wrote an article that contained only the elements I listed as "certainly secondary" and submitted it to a science journal as a research paper, it would surely be rejected for lack of originality, wouldn't it? --Gerry Ashton (talk) 19:04, 17 December 2007 (UTC)
Do you mean "certainly secondary"? Anyway, my point is not that your placement of the material into the system is wrong, but that scientists would not recognize the way we use the terminology -- they would call a paper "primary" that according to your scheme would be called "secondary". (Although if I am not mistaken SlimVirgin has described that kind of peer-reviewed paper as "primary", but meaning something different than we mean; meaning that it should be disfavored since it is a description by the scientist of the scientist's own research, hence "close to the source".) Since the usage is inconsistent editors coming from science disciplines will be confused. --Lquilter (talk) 20:01, 17 December 2007 (UTC)
The meaning of primariness and secondariness is set forth very well in the primary source and secondary source articles. Primary and secondary are relative terms, and like any relative terms (e.g., up/down, right/left, earlier/later), they depend on context. You can't say a source is always primary or always secondary, because it depends on how it is cited. There are certain types of works that are almost always cited in a primary way, such as journal articles, interviews, diaries, and raw data, but these can just as easily be secondary too, like if you cite a scientific journal for its comments on another scientific journal. COGDEN 21:06, 17 December 2007 (UTC)
What you say is true. Usage depends on context and on the field/discipline using the sources. It is also true that secondary sources can be biased, and that great care must be taken when using them, perhaps more so than primary sources. Sources must also be evaluated for fairness and objectivity. The current policy ignores the most important issues, which I find strange. —Viriditas | Talk 23:58, 17 December 2007 (UTC)

Proposal to shut up

Since my proposal of five and a half hours ago to summarize the current situation because there was too much going on on the talk page there have been five new sections added to this talk page, meaning that it's expanding at roughly one per hour.

In this much of a crush of "me-too" comments and disparate proposals it is not meaningfully possible that this dispute is going to resolve in a productive manner.

I ask again - what are the actual areas of dispute here? Please. Quote the wording in the current or a recently proposed version of the policy that you dislike. State your objection. Ideally, propose alternate wording that you would support. Let's try to keep this to actual discussions on content and see if we can come to agreements on wording.

Or, I mean, we can all introduce differing proposals and say "me too" a lot. But I'd rather see if we can actually do something with the page instead of the talk page. Phil Sandifer (talk) 23:13, 17 December 2007 (UTC)

The proposal is to bring this and every other policy and guideline in line with Wikipedia:Simplified ruleset, making usability issues paramount above unneeded complexity. —Viriditas | Talk 23:22, 17 December 2007 (UTC)
I think we're making progress. Jossi has proposed opening up a subpage on source categories, and Brimba's written a tightened lead. We're moving forward at last. SlimVirgin (talk)(contribs) 23:23, 17 December 2007 (UTC)
I would like to know why it is acceptable to state that "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" but unacceptable for Vassyana to add "secondary sources are usually preferable because they generally provide analysis and a broader context for the subject, and usually offer a more independent view. However, they can be biased by the views of their author(s) and/or publisher, so caution is required to preserve the neutral point of view. Care should also be taken to avoid undue weight and ensure the information cited is used in context." This is a very important point, as secondary sources can be misused more than primary sources on Wikipedia as they are our preferred method of sourcing to represent other POV. —Viriditas | Talk 23:52, 17 December 2007 (UTC)
Godo points. Please help working out the wording for the NPOV section at the Wikipedia:Source typing sandbox. ≈ jossi ≈ (talk) 00:35, 18 December 2007 (UTC)
A quick answer is to read the talk page and compare the edit/revert history. That will show why one was okay and the other wasn't. Not much more needs to be said. wbfergus Talk 00:51, 18 December 2007 (UTC)
That shows nothing. Vassyana's modifications were important and relevant; there was no reason stated for their removal that I can see. —Viriditas | Talk 01:27, 18 December 2007 (UTC)

Re: Phil Sandifer' request for identification of the areas of dispute. These, according to the third eye :p, are:

a) the relevance of a distinction between "primary" and "secondary" sources with respect to "original research";
b) the applicability of the terms "primary" and "secondary", to include the arbitrariness of such definitions, and the problems that such definitions cause;
c) the pitiful prose in that section.

To which the kind Saint D. <g> said:

I believe this is a very succinct and accurate summary of the questions that have been raised on this talk page that still remain unaddressed, and should be discussed further here.

-- Fullstop (talk) 01:59, 18 December 2007 (UTC)

NotabilitySignificance is central to OR?

Apologies if this conflicts with something already going on in the last 5 sections. I've been distracted framing what follows. Here goes.

Background/Introduction
In this comment SlimVirgin asserted that "notability is central to OR."
Now, I've seen this "Notability is central to OR" mantra repeated ad-infinitum, with no one - until Dave souza - having explained how this might be so. (Thanks Dave!)
In this comment, Dave notes that "notability" is not in fact confined to WP:Notability, but is rather "in the general sense of meaning 'significance in relation to the subject'." With this, Dave has proffered a glimpse of what the heck that mantra is supposed to mean. It however leaves some other issues I'd like to address.
A word of caution in advance though: none of these is new. Although my conclusion will come back to this note, please do not gloss to the end.

--

'Significance' and NOR
  • Even in the most general sense of the idea of 'significance' of a source, there is no relationship to OR. Even the most significant of sources can be subjected to OR. Inversely, even the least significant of sources can be subjected to OR. I know of at least one article that consists entirely of misrepresented secondary sources. The ostensible subject/topic of the article has been academically referred to, the sources are themselves 'significant', but almost every sentence in the article is still a complete fraud, down to the faked interwikis.
  • But how on earth can significance be central to OR anyway? The ability to corrupt a source is not influenced by the significance of that source. Its utopian to assume that editors are going to be thinking "out of respect for the source's significance I shall not misrepresent it" or even "let me find a less significant source to upchuck onto wikipedia."
  • The significance of a source is not something that can be influenced by NOR policy either. Not only do we have WP:RS policy to govern the suitability/significance of a source, the 'process' that leads to misrepresentation of a source takes place after the source has already been selected.
The 'significant' source, once chosen (ideally in accord with WP:RS), is secondarily going to have statements taken from it. Whether those statements are going to be accurately represented on WP, or whether they are going to be warped to say something that they do not, is a process independent of how 'significant' that source is.
In fact, editors who are out to push a stance will intentionally look for a source that (vaguely) "reflects" what they want to push, the more 'significant' the source turns out to be, the better. I have seen this happen myself. Not 'just' with secondary and tertiary sources, but even with the finest authoritative sources.
Under those circumstances, it is meaningless to invoke NOR. PSTS in NOR effectively cripples any invocation of NOR. One can't say "Hey, thats OR!" The editor will simply turn around and say "No, thats not. This is the finest source there is." Yes, of course it is gaming the system, but when was the last time that an admin took the trouble to explain things to a recalcitrant editor? Its much too much trouble to go to Arb/Mediation over content, even if an editor could be sure of the outcome. Given that pure content disputes are almost impossible to resolve by an (essentially uninterested) third-party, an editor would have to force a war to find something actionable first. So the smart editor shrugs his shoulders and goes away. The crud stays.

--

Conclusion
Whatever meaning 'notability' is supposed to have, the notion that it is "central" to 'no original research' policy, or even that there is any correlation at all, is far out of touch with how sources can be and are misused. Its based on the utopian premise that editors are not going to be using everything in their power to misuse WP. If this were so, and sadly its not, we wouldn't need policies.
Now, as I noted in the introduction, none of these points is new. The typical response to any of them has thus far been "oh, the policy is fine. The editor is at fault." That is not good enough! Non-administrators have no leverage.
Editors have no active protection in a content dispute. And we desperately need it. If not by admins directly, then at least in being able to invoke policy. To know that what we are doing is good and has blessing is the least we should be able to expect.

Sorry if this has turned into an essay in its own right. I hope its not too boring to read. :) -- Fullstop (talk) 01:29, 18 December 2007 (UTC)

In the sandbox Wikipedia:Source typing there is a section about Notability, in which you expand on your points above. ≈ jossi ≈ (talk) 01:50, 18 December 2007 (UTC)
I agree strongly with this - WP:N is already one of our most problematic concepts, and is the source of a terrifying share of our bad press. No small portion of the problem is that WP:N is based on excessively programmatic and rigid perspectives on sourcing that lead towards arbitrary conclusions. I strongly oppose using WP:N as any foundational part of this policy simply because this is a policy, and WP:N is an ugly hodge-podge that continues to need a heavy cleaning up. Phil Sandifer (talk) 02:05, 18 December 2007 (UTC)
I also agree that some of the things this part of the policy is trying to articulate might better be expressed as statements about notability and its sourcing. Statements about notability might be better off in WP:N then WP:OR. Best, --Shirahadasha (talk) 02:55, 18 December 2007 (UTC)

Reverts

Vassyana, you made significant changes with no consensus, yet when other people undo some of them to restore old language, you revert saying they need to discuss first on talk. But it has been discussed. There is no consensus for those changes, so please don't keep introducing them. Some of what you changed did stay, but some parts were very problematic e.g. removing that articles should be based on secondary sources. SlimVirgin (talk)(contribs) 17:13, 17 December 2007 (UTC)

All the changes I've made were based on discussion and well-advertised on the village pump and RfC. You do not get to mystically assert "consensus" or "no consensus" or push your opinion through edits any more than anyone else. You are doing exactly the sort of editing that you criticize COGDEN for. If you disagree with changes to the policy you are free to contribute to the discussions like anyone else. Vassyana (talk) 17:17, 17 December 2007 (UTC)
I should also add that asserting your edits have been proposed or discussed is a bald-faced lie simply not true. Vassyana (talk) 17:18, 17 December 2007 (UTC)
Gals, guys, please take a deep breath... This discussion is becoming a bit over the top. ≈ jossi ≈ (talk) 17:24, 17 December 2007 (UTC)
It's becoming beyond ridiculous.[29] I more than a little tired of being told by <expletives> that haven't bothered to discuss their changes to go to this talk page to discuss the changes that we've already discussed and agreed upon!!! So, pardon me if my comments are mildly over-the-top, because that kind of asinine behavior deserves far less kind language. Vassyana (talk) 17:38, 17 December 2007 (UTC)
Vassyana, you're going way too far. SlimVirgin (talk)(contribs) 17:26, 17 December 2007 (UTC)
Hardly, unless you care to point out where I'm crossing the line. I don't see calling a spade for a spade going anywhere near too far. The simple facts are that the changes I made were well-discussed and revised according to feedback, while your edits have not been proposed and discussed (quite contrary to your assertion). You're perfectly welcome to join the discussion, if you would care to. Vassyana (talk) 17:38, 17 December 2007 (UTC)
Vassana, your behavior here is completely unacceptable. To call your fellow editor "a bald faced liar" is a personal attack. If you persist with this type of behavior, you will be blocked per WP:NPA. If you want to effect changes, you must reach consensus with everyone, not just a sub-group, and you cannot attack those who disagree with you as 'liars'. Crum375 (talk) 17:48, 17 December 2007 (UTC)
Take your own advice about reaching consensus. Why not propose your edits like everyone is required to? If some people will not express their opposition in a reasonable fashion, there is no obligation to accept their dissent. It's not a matter of attacking someone who disagrees with me and if you believe so, I guess you've your opinion. Vassyana (talk) 17:56, 17 December 2007 (UTC)
I don't see Vassyana's comments as crossing the line or going to far. I see it as venting frustration with dealing with what appears to many people (judging by the comments here and other places) of a double-standard. That's all. On one hand the original (now stricken) statement could possibly be interpreted as a personal attack (similar statements in the Old West did generally lead to a gunfight), it was more or less a statement about one specific incident, not a gross-overstatement regarding all of Vassyana's opinions about anything and everything dealing with SlimVirgin. wbfergus Talk 14:12, 18 December 2007 (UTC)

Would someone be so kind as to summarize the remaining issues here for me, or for anybody else whose eyes glazed over 10 sections ago? Phil Sandifer (talk) 17:33, 17 December 2007 (UTC)

Me too. --Lquilter (talk) 17:35, 17 December 2007 (UTC)
And me three. The talk page is an unwieldy 500K in length, with 300 edits just since this weekend. Plus another 40ish to the mainspace. At this point, anybody rushing to edit that main space, I don't care if it's supposedly a consensus edit, a "there is no consensus" revert, or to simply to make a spelling correction!, is just adding to the madness. I say we archive the whole page and take some kind of a break. With a fresh start, we agree to follow something at least roughly resembling a systematic approach to finish this, by setting a timeframe target for one thing, so editors can actually follow along and contribute to the beginning, middle, and end to this mess and get on with actually editing articles around here. Professor marginalia (talk) 20:34, 17 December 2007 (UTC)
I concur with Prof M. After all these years er, months of struggle, are we rushing to get changes into place before Christmas holiday? ;-) Dreadstar † 20:43, 17 December 2007 (UTC)
All I want for Christmas is a NOR policy that every single Wikipedian can heartily endorse as clear, accurate, concise and perfectly understandable to every person with a high school education. Is that asking to much?     :^)      wbfergus Talk 20:52, 17 December 2007 (UTC)
I think you're more likely to get that ipod. <g> --Lquilter (talk) 21:25, 17 December 2007 (UTC)

Life beyond PSTS -bloat reduction

Rarely has any policy page had as many experienced editors looking over it as this page has at this time. While all, or most, of the attention has been on PSTS, the page overall is in bad shape, suffering from extensive bloat, and poor prose. For example the lead is far larger than WP:V’s lead, and adds very little addition functionability. At the other end of the article, the section “Other options” could be replaces with a single well written sentence. Any legitimate opinions as to how to return this page to a K.I.S.S. format would be welcome. A lot of stuff in between could use a good copyedit, or outright pruning.

The KISS principle states that design simplicity should be a key goal and unnecessary complexity avoided. Complexity should, very simply, be avoided: simplicity becomes a goal in itself. Extra features are not needed; an approach that seems "too easy to be true" is in fact the best way; a very straightforward approach may seem less glamorous and less dramatic but that trivial approach should indeed be taken; surprisingly the very obvious approach is often best. The principle is somewhat similar to…Albert Einstein's maxim that "everything should be made as simple as possible, but no simpler." KISS principle

Brimba (talk) 19:44, 17 December 2007 (UTC)

Good idea - I'm a big fan of KISS. For what it's worth I've never understood the example in the Synthesis section. It almost got replaced with something comprehensible a while ago, but, of course the change was blocked. Good here isn't it!  —SMALLJIM  20:05, 17 December 2007 (UTC)
(ec), Brimba, the page was expanded a few months ago for reasons that were unclear, because it was better before. The more people get involved in editing it, the longer it gets. It does need a good copy edit to tighten it. SlimVirgin (talk)(contribs) 20:06, 17 December 2007 (UTC)
SV, can you provide a diff for the version you think was "better before"? --Lquilter (talk) 20:09, 17 December 2007 (UTC)
Looking at the lead, this is what it used to say:

Wikipedia is not the place for original research. Citing sources and avoiding original research are inextricably linked: the only way to demonstrate that you are not doing original research is to cite reliable sources that provide information directly related to the topic of the article, and to adhere to what those sources say.

No original research is one of Wikipedia's three core content policies. The other two are Neutral point of view and Verifiability. Jointly, these policies determine the type and quality of material that is acceptable in the main namespace. Because the three policies are complementary, they should not be interpreted in isolation from one another, and editors should try to familiarize themselves with all three.

SlimVirgin (talk)(contribs) 20:14, 17 December 2007 (UTC)
This version from Sept. 30 (diff) hews pretty closely to that, with one additional paragraph at the top; or is there another one that is closer? I agree that the language you quote above seems cleaner, simpler, and easier to read than the current language (12/17 diff), although there are some parts of the 12/17 version that are nice; a merge that followed the 9/30 version generally with a little bit of the phrasing from 12/17 could be very good. --Lquilter (talk) 20:31, 17 December 2007 (UTC)
While this conversation was going on, I reworked the lead to read as follows (just a suggestion), however it looks fairly close to what existed in the past.
Wikipedia does not publish original research or original thought. This includes unpublished facts, arguments, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position. This means that Wikipedia is not the place to publish your opinions, experiences, or arguments. Citing sources and avoiding original research are inextricably linked: the only way to demonstrate that you are not presenting original research is to cite reliable sources that provide information directly related to the topic of the article, and to adhere to what those sources say.
Wikipedia:No original research (NOR) is one of three content policies. The others are Wikipedia:Neutral point of view (NPOV) and Wikipedia:Verifiability (V). Jointly, these policies determine the type and quality of material that is acceptable in articles. Because they complement each other, they should not be interpreted in isolation from one another, and editors should try to familiarize themselves with all three.
Brimba (talk) 20:50, 17 December 2007 (UTC)
Very good. I'd support that to replace the current lead. SlimVirgin (talk)(contribs) 20:52, 17 December 2007 (UTC)
Most excellent, Brimba. Dreadstar † 20:54, 17 December 2007 (UTC)
Yes, good job! Three minor suggestions: (1) Instead of "readers should try to familiarize" why not "readers should familiarize"; try is rather implied. I would also suggest that the same language be placed on all three policies for consistency.
(2) Drawing from one of the turns of phrase I liked from the 12/17 version, I propose "... and to represent accurately what those sources say." Currently it says "and to adhere to what those sources say." Adherence has a flavor of "follow closely", as in follow a guideline; whereas we're talking about paraphrasing or describing closely, not following as in obedience. So "represent ... accurately" I think is a bit better and also simpler language than "adherence".
(3) Can make it even simpler and stronger and more active-voice by striking "the only way".
Thus:

Wikipedia does not publish original research or original thought. This includes unpublished facts, arguments, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position. This means that Wikipedia is not the place to publish your opinions, experiences, or arguments. Citing sources and avoiding original research are inextricably linked: the only way to demonstrate that you are not presenting original research, you must is to cite reliable sources that provide information directly related to the topic of the article, and represent those sources accurately. to adhere to what those sources say.

Wikipedia:No original research (NOR) is one of three content policies. The others are Wikipedia:Neutral point of view (NPOV) and Wikipedia:Verifiability (V). Jointly, these policies determine the type and quality of material that is acceptable in articles. Because they complement each other, they should not be interpreted in isolation from one another, and editors should try to familiarize themselves with all three.

--Lquilter (talk) 21:08, 17 December 2007 (UTC)


query on "to advance a position"

Are there syntheses, previously unpublished, which do not advance a position and which Wikipedia would therefore publish? In other words, is "to advance a position" a superfluous phrase that could be deleted? Or, is it helpful because any article written is, to some extent, a previously unpublished synthesis, albeit only in the sense that it is a new tertiary source; obviously it must not advance positions or novel ideas. ? --Lquilter (talk) 21:19, 17 December 2007 (UTC)

No, not superfluous but crucial. All Wikipedia articles are previously unpublished syntheses. What's not allowed is for a position to be advanced via synthesis. SlimVirgin (talk)(contribs) 21:22, 17 December 2007 (UTC)
Fine. I just wasn't sure if that was the intended use or not -- if so, then as you say it's essential to ensure that people don't start demanding we only reprint previously published articles. <g> --Lquilter (talk) 21:24, 17 December 2007 (UTC)
I have no problems with Lquilter’s main changes, although I would retain the phrase “The only way”. I too think we should retain “to advance a position”. Brimba (talk) 21:33, 17 December 2007 (UTC)
That doesn't work the "you must" construction; do you want to get rid of that as well? (Can you post what you're thinking of that combines my comments that are okay with you?) --Lquilter (talk) 21:44, 17 December 2007 (UTC)
Ummm..scratch that. Looks fine now that I see the entire thing. Brimba (talk) 21:55, 17 December 2007 (UTC)
Looks fine to me. Go for it. :-) SlimVirgin (talk)(contribs) 23:25, 17 December 2007 (UTC)
Actually to avoid the revert wars I would suggest holding off on "going for it" just for a little while longer (a day or two) to seek more consensus. --Lquilter (talk) 23:42, 17 December 2007 (UTC)
Looks fine. Dreadstar † 07:43, 18 December 2007 (UTC)

Nutshell

Simple question, would there be any objection to condensing the nutshell as so:

This version seems both clearer, and more on point. Brimba (talk) 01:05, 18 December 2007 (UTC)

Excellent. -- Fullstop (talk) 01:36, 18 December 2007 (UTC)
ps: Since you've initiated an overhaul of prose, would you please also assign tasks so we don't end up with multiple people working on the same sections? I volunteer to tackle anything. :)
  • The third point is a little unclear to me. Maybe use the language from the preface: "Articles should not include unpublished analysis or syntheses of published material that serves to advance a position." It's a little longer, though. --Lquilter (talk) 01:50, 18 December 2007 (UTC)
Thanks for the endorsement, but I would be a little uncomfortable assigning tasks. I think the only rule is to avoid editing PSTS (to keep this project separate from the ongoing edit war).
If we include “Articles should not include unpublished analysis or syntheses of published material that serves to advance a position.” Point three largely duplicates point two. However we could go with this:
Point two comes from WP:ATT, which conveys the idea clearly. Brimba (talk) 02:09, 18 December 2007 (UTC)
  1. Point #2 is trespassing into RS. Don't. :)
  2. Break down the run-on in point #3 into
    • Articles should not include unpublished analysis.
    • Articles should not include syntheses of published material. (link WP:SYN here)
  3. Instead of 3a (above), how about "Articles should not misrepresent a source."
    (or something like that to cover all OR evils, i.e. "creative" interpretation, out-of-context OR, synthesis etc)
  4. Also, why the "should not" and not "may not"?
-- Fullstop (talk) 03:01, 18 December 2007 (UTC)
  1. WP:RS is simply a guideline that explains WP:V. This page on the other hand interlocks directly into V. You can not understand or implement V in isolation from NOR, or the reverse. Stating that relationship upfront is basic to stating what NOR is about, and conveying to the reader what NOR is about. To reserve something for a guideline such as RS is to force people to look in two places for information that interconnects; I don’t get the point of your objection.
  2. I would rather have one run-on sentence under one bullet, than to replace that with two separate bullets. And doing so would drop the phrase “that serves to advance a position.” Language is about conveying information, if a run-on sentence is able to covey ideas accurately in a concise manner, then I can live with that.
  3. “Articles should not misrepresent a source.” Is not a bad idea; however, the Nutshell should stand alone, as should the lead. People combining A+B so that they = C may not equate what they are doing with misrepresenting a source. I think it is much better to state that directly. Many will see “misrepresenting a source” as a prohibition against editing in bad faith, and fail to see that their own good faith efforts to improve articles violate policy.
  4. "may not" is probably an improvement. Sounds good to me. Brimba (talk) 05:49, 18 December 2007 (UTC)
So what's the latest version? FWIW I liked the last orange box by Brimba. --Lquilter (talk) 08:06, 18 December 2007 (UTC)
Brimba, yes, I see what you mean. Per #3, I actually meant 3b, not 3a. Sorry for the confusion.
And heck yes, I can only hope that "Many will see “misrepresenting a source” as a prohibition against editing in bad faith"! :)
Taking your points into consideration (and fixing the run-on in another way), the points would read:
  • Wikipedia does not publish original research or original thought.
  • All material in Wikipedia must be attributable to a reliable, published source.
  • Articles may not include unpublished analysis, and may not misrepresent sources in order to advance an editor's position.
I've highlighted only the phrases that differ from your last suggested nutshell.
-- Fullstop (talk) 17:26, 18 December 2007 (UTC)

Proposal to move forward

Would editors consider useful to explore the creation of a PSTS page in a sandbox?

The idea could be:

  1. Create a PSTS page in which the subject of primary, secondary, and tertiary sources is explored in general
  2. The PSTS page will have specific sections about the application of source typing to WP:NOR, WP:NPOV, WP:V, and any other pertinent policy
  3. Create summaries of these sections that can be featured in the main policy pages as per WP:SUMMARY

I think that pursuing this will be more productive that being stuck in this debate here. ≈ jossi ≈ (talk) 22:03, 17 December 2007 (UTC)

  • I would support this. (Of course, I'm supporting just about anything that moves this <g> ... but this is still a good idea.) --Lquilter (talk) 22:28, 17 December 2007 (UTC)
  • It's not often that Jossi and I are on the same page, but I think this is a good idea. We tried to do something like this a while ago in an essay (Wikipedia:Classification_of_sources, now probably out of date), but it was quickly forgotten like a lot of other proposals. The only thing I would want to ensure, however, is that everyone understands the sandbox page is just scratch paper for working out the issues, and for getting everybody's concerns out in the open. If there are multiple points of view as to a particular issue, instead of edit warring, we can be inclusive (as if this were a content article including multiple POVs) so that we can see how different POVs compare, and then people should feel free to try and reconcile them. I think that if we put everything on the table in one place, we'll see that most of our arguments are just semantics, and that there is a lot we can fundamentally agree on. The issues that we really do disagree on can then be more readily apparent. COGDEN 00:11, 18 December 2007 (UTC)
Started the sandbox at Wikipedia:Source typing. Happy editing! ≈ jossi ≈ (talk) 00:22, 18 December 2007 (UTC)
Just a general 'placement' question. Any particular reason the page was created there instead of in the 'Sandbox' for this policy, per the link near the top of the page at (I beleive) Slrubenstein's suggestion several months ago? Any difficulties in moving it? Thanks. wbfergus Talk 14:19, 18 December 2007 (UTC)

Towards a compromise: A modest proposal

First, some premises on which the proposal is based:

  1. The current primary/secondary/tertiary approach has the problem of saying too much (overspecifying), and as a result sometimes leads to counterintuitive results.
  2. However, saying nothing at all would be worse than the present approach, because not all sources are equally usable for all types of claims, and we need to say something that addresses this.
  3. The proposed compromise is to say less, covering only what's required to make the policy make sense and no more.
  4. To do this, I propose simply stating the necessary requirements in policyspeak, using terms like notability, significance, connection, independence, and expertise. Doing so lets us specify only what is needed for core policy/guideline purposes, without classifying sources for any other purpose so there is no need to impose a general classification system on sources that might lead to problems.
  5. This first attempt to do this is very likely wrong, but hopefully can be improved into something more workable.

Here is what I understand to be the key principles that the PSTS system is intended to protect

  • Principle: Notability and significance require independent sources
    • Sources who are directly connected with a topic (e.g. invented/discovered, advocate, work for, etc.) should not be used to assess the notability of that topic. Notability should be assessed by independent sources.
    • Sources who are directly connected with a point of view should not be used to assess the significance of that point of view. Significance should also be assessed by independent sources.
  • Principle: Involved sources can be used once notability/significance has been independently established
    • Once notability has been established by independent sources, involved sources can be used for general facts and opinions about the topic.
    • Once significance has been established by independent sources, the topic's advocates can be sourced to articulate it.
  • Principle: independence/expertise tradeoff
    • In fields requiring scholarly expertise, works by experts in the field are preferred to non-expert sources such as general media.
    • However, if an independent expert is not available, notability and significance can be determined by independent general media sources.

It seems to me that the purpose of bringing in the primary/secondary/tertiary concepts into the policy pages is a belief that these phrases articulate concepts like notability, significance, independence, involvement, expertise, etc. I believe these concepts would be better expressed directly. I also believe WP:SYNTH might best address analytic and synthetic claims. Best, --Shirahadasha (talk) 02:39, 18 December 2007 (UTC)

Note: This very likely isn't all the things the policy needs to cover. It's only a first stab. A more complete set of "policyspeak" statements is likely needed. Experienced editors might better understand "policyspeak" statements, and might be better able to negotiate about them, than statement expressed in the current framework. Best, --Shirahadasha (talk) 02:52, 18 December 2007 (UTC)

Shirahadasha, Could you tie all of that to the concept of "No Original Research". At the moment your proposal does not seem to say anything about how it relates to the policy in question. Blueboar (talk) 03:44, 18 December 2007 (UTC)
My personal view is much of the detail on types of sourcing for different purposes doesn't belong in WP:OR -- I think it belongs in WP:N etc. I believe rephrasing in policyspeak tends to make this clearer. Example, when a "primary" source is used for a POV, I agree the source is reliable for the viewpoint itself, but not its significance. By rephrasing the problem in policyspeak, I see it as a WP:SIGNIF problem, not a WP:OR problem.) However, others appear to disagree with me, and I may not understand their viewpoint adequately. It's quite possible this synopsis omits requirements which perhaps more clearly do belong in WP:OR. Best, --Shirahadasha (talk) 04:23, 18 December 2007 (UTC)
Among the many problems with WP:N, I don't believe that source typing has been one of them. The relevance has been questioned, but no one has tied themselves in knots about the set of possible sources that may be either primary or secondary. In a theoretical sense, source typing seems confusing, but on a case by case basis it is straightforward. Lets just leave the definitions, nuances and alleged contradictions to the mainspace articles where debate is limited to verifiable reliable sources. The theory of source typing doesn't belong in WP:NOR, and it doesn't belong anywhere else in project space either. --SmokeyJoe (talk) 11:33, 18 December 2007 (UTC)
Shirahadasha says "* Sources who are directly connected with a topic (e.g. invented/discovered, advocate, work for, etc.) should not be used to assess the notability of that topic. Notability should be assessed by independent sources." I think this language is a bit confusing -- for instance, one might read it and say "a peer reviewed article is assessed by independent sources". We can rephrase this to avoid the confusion and get at what I believe we all understand to be the core -- that whatever the concept or idea is that is the subject of the article, independent recognition must be made of it. Possible wording might be "Any topic should be notable. Multiple independent published sources that directly describe the subject as notable directly establish notability. Multiple independent published sources on the topic of the subject provide indirect evidence of its notability, whether they describe it as 'notable' or not." --Lquilter (talk) 05:59, 18 December 2007 (UTC)

secondary sources II

(Outdent) I see in the 'policy of the hour' what seems to be a glaring problem with the current definition for 'Secondary sources'. The policy currently states "They draw on primary sources to create a general overview; or to make analytic or synthetic claims", which is an incorrect (or incomplete) analysis of the two refs provided. Can "primary" simply be replaced with "other"? That makes the sentence more accurate (many secondary sources are also based entirely on other secondary sources). wbfergus Talk 21:25, 17 December 2007 (UTC)

Most secondary sources, by definition, draw upon primary sources. They may use other sources that are not primary, though; so we could say; They draw mainly on primary sources to create a general overview; or to make analytic or synthetic claims ≈ jossi ≈ (talk) 21:31, 17 December 2007 (UTC)
There is a purposeful flow through all three source statements that shows the interrelation between them. Keeping "primary sources" as part of the statement on seconday and tertiary is necessary in order to help understand their relationships to each other. Secondary sources draw from primary and other sources, but if we include "other sources", we would need to nail down what that means...e.g. do secondary sources draw from tertiary sources? What are these other sources? Things like that. Jossi's version is good. Dreadstar † 21:37, 17 December 2007 (UTC)
My office within USGS complies all sorts of data from only secondary sources, and the results are a combination of primary and secondary material. They (the folks in my office) rarely, if ever, use primary source data. Our parent group however, uses only primary data to produce secondary data. Last time I was in our library a few weeks ago, I was reading through all sorts of geology related books which all almost entirely, if not completely based upon other secondary material. The authors had merely gathered together other seconadry source information and regurgiated it into another variation that was primarily secondary, but also drew new analysis or other new 'ideas' previously unpublished, so for those particular conclusions, they were also primary sources. The way the sentence currently reads, all secondary sources are based entirely, and only, on primary sources. That is very incorrect. I wouldn't even say that secondary sources draw mainly on secondary sources, as I can find thousands of books in our library quite the opposite, not to mention the thousands of pubs our organization churns out. wbfergus Talk 00:42, 18 December 2007 (UTC)
Perhaps we are once again up against the difference between how the terms are used in the humanities and how they are used in the sciences? Jossi's version certainly fits how the term is used in the study of History. Blueboar (talk) 01:35, 18 December 2007 (UTC)
No, not necessarily. A source is a source is a source. It can - subject to the context for which it is being cited - be either primary (if the statement using that source refers to to the source for something original to that source) or secondary (for the analysis of raw data) or even tertiary (because a good secondary source would also acknowledge/debunk previous research). This can also happen for any source in any field if the data and its analysis appeared in in one and the same text. An archeology/sociology text or a physics/biology text wouldn't be that different in this respect.
The distinction is then between field research where new data is being collected and analyzed, and ivory tower research which just addresses preceding data and theories.
-- Fullstop (talk) 06:00, 18 December 2007 (UTC)
WRT humanities/sciences, there are some things in Wikipedia:Source typing#In the humanities & Wikipedia:Source typing#In the sciences that need looking at by as many eyeballs as possible. Even though I just wrote the "In the humanities", I'm already inclined to write it off as crud because I can already see ways in which its "all" wrong.
-- Fullstop (talk) 06:00, 18 December 2007 (UTC)
So is there agreement to change the wording of that sentence (and only that sentence), with replacing the word 'primary' with 'other'? The supplied cites don't even agree with the sentence (which in an article would lead to all kinds of problems). The closest cite states "Secondary sources are works of synthesis and interpretation based upon primary sources and the work of other authors". This has the extra inclusion of 'other authers', while the policy in the sentence reads that it is only from primary sources. So, simply replacing 'primary' with 'other' clearly matches the supplied citation, is far more accurate, and elimates a false assertion. Agreed, and does anyone feel like making a "bold edit"? wbfergus Talk 14:06, 18 December 2007 (UTC)
Well, if PS/SS distinction - being irrelevant to NOR - goes elsewhere, then its kinda moot. If it doesn't then "They draw on primaryother sources to create a general overview; or to make analytic or synthetic claims" is fine and more correct and non-OR :) -- Fullstop (talk) 16:17, 18 December 2007 (UTC)
I don't there's an agreement to replace "primary" with "other", I was just questioning that very thing in my post above, wanting to know what exactly "other" meant, and I even put forth "Secondary sources draw from primary and other sources", if we were able to define what "other" actually meant. No replacement agreement yet. And now, <sigh>, I need to review what Francis just did...a truly bold edit! Oh, wait..it's been reverted...well...here we go....lol... Dreadstar † 20:33, 18 December 2007 (UTC)
Oops, then my 'bold' edit will get lost i n the shuffle. The supplied citation for the sentence states 'other authors', but with that change, then the sentence itself is almost an exact copy of the cite. So, I thought the 'authors' would simply be implied, but I guess in some situations (lack of coffee, lack of sleep, etc.) it could be interpreted ambiguously as well. so maybe 'authors' would be needed also. wbfergus Talk 21:35, 18 December 2007 (UTC)
It was an excellently done Be bold! edit, so no oops! I'm not sure what "other" means when it's referring to primary, secondary and tertiary sources..how is other being defined an how do we make sure it's understood by anyone reading the policy? And yeah, the recent reverts were quite surprising and shuffling..;) This is quite the active policy! Dreadstar † 21:41, 18 December 2007 (UTC)
I found Wbfergus' edit here to be reasonable. Understanding that it's tough to cover all conceptual bases with a single swipe, I imagine the sentence might reasonably say something like the following? "Secondary sources draw upon primary sources, or upon other secondary sources, to make analytic or synthetic claims, or to create a general overview." ... Kenosis (talk) 21:55, 18 December 2007 (UTC)
That's a good statement, Kenosis. I reverted wb's edit because it removed "primary", and I think it's important to keep "primary" in the three source type descriptions to help provide a better understanding of their relationships to each other. My real question was what are other sources, if all others are just other secondary sources, then your proposed statement is great! Dreadstar † 08:26, 19 December 2007 (UTC)
I agree with Dreadstar. That does help clarify the sentence better, more accurately reflects the citatation, and eliminates the implied reliance of secondary sources solely upon primary sources. Thanks guys. wbfergus Talk 11:26, 19 December 2007 (UTC)
  1. ^ Borough of Manhattan Commmunity College, A. Philip Randolph Memorial Library, "Research Help:Primary vs. Secondary Sources" notes that a secondary source "analyzes and interprets primary sources", is a "second-hand account of an historical event" or "interprets creative work". It also states that a secondary source "analyzes and interprets research results" or "analyzes and interprets scientific discoveries".
  2. ^ The National History Day website states simply that: "Secondary sources are works of synthesis and interpretation based upon primary sources and the work of other authors."
  3. ^ University of California, Berkeley library
  4. ^ Hacker, Diana. (1999). A Writer's Reference. Fourth Ed. p. 70. ISBN 0312260377
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:No_original_research/Archive_31&oldid=1142664225"