Wikipedia talk:No original research - Wikipedia


1 person in discussion

Article Images

Template:NORtalk

Please join a discussion at Wikipedia talk:Verifiability#Text vs. other media. Maurreen (talk) 16:05, 3 March 2010 (UTC)Reply

I wonder if we need stronger wording regarding synthesis by "phrase matching" of a neologism. I keep seeing articles survive AfD because they are titled a string of words that occurs in reliable sources, often in passing. These passing occurrences of the words are then cobbled together to write an article on the synthesized "concept". Examples are articles like Post disco, which primarily rely on sources mentioning the title of the article (or something like it) in a very offhand way. Many musical genre articles are using this kind of "phrase synthesis". Take an existing genre and stick post- on it as a prefix, and you'll find enough hits to cobble together what looks like notability (and many of these post-genre articles actually do exist). Another example is a currently running AfD that I recently posted which I will not link to in order to avoid canvassing.

I think our guidelines fail here because notability assumes that the subject actually exists in a meaningful sense, and if it doesn't, there won't be coverage. But when it comes to concepts, subjects themselves can be synthesized. WP:NEO says that sources should be about the word rather than merely using the word, but that's the closest we have to a useful guideline on this issue. As part of the MOS and not a policy or guideline, it seems that the advice in WP:NEO is not taken very seriously at deletion discussions. Gigs (talk) 21:03, 10 March 2010 (UTC)Reply

I agree with you, but you're asking for a much bigger thing than you realize. you want issues like this to be decided by due consideration and reasoned discussion, but the AfD system is only designed to support a loose form of voting, and often gets swamped by tendentious editors trying to back up their original SYN with more protracted forms on syn. I don't know how to fix that, I'm just pointing out the problem. --Ludwigs2 21:30, 10 March 2010 (UTC)Reply
I agree it is a tough nut. It cross cuts notability, OR, proper use of sources, and our definition of consensus that lets one or two editors foul an AfD to no consensus on very weak policy grounds. Surely we can come up with something though, right? Gigs (talk) 00:42, 11 March 2010 (UTC)Reply
I agree completely... not sure if this is the best policy to discuss it in (it is related to NOR, but only tangentially)... but we do need to discuss neologistic phrases somewhere. Blueboar (talk) 01:03, 11 March 2010 (UTC)Reply
I'm thinking that the notability angle might be the better place for it. If we made it more clear that sources merely using a phrase does not constitute "coverage" of that phrase then we could probably dodge the issues of original research that are often part of such articles. Gigs (talk) 19:17, 11 March 2010 (UTC)Reply
Agree... discuss it at WT:NOTE and I will support. Blueboar (talk) 19:19, 11 March 2010 (UTC)Reply

Is a nonsense statement. It is neither true, not based on the reference offered. A secondary source is a first hand account of the author of that source in saying something about something else. Suggesting that "second hand accounts" are by definition "secondary source" is wrong. "Second hand accounts" sounds more like unreliable repetition. --SmokeyJoe (talk) 02:46, 11 March 2010 (UTC)Reply

It was correct as written. SlimVirgin TALK contribs 02:49, 11 March 2010 (UTC)Reply
OK, maybe it can be read correctly, when read in full, but our audience is prone to stopping early, leaving an erroneous message. It is not useful to start up with "are second-hand accounts", as this is not a defining statement, and is problematic. --SmokeyJoe (talk) 02:54, 11 March 2010 (UTC)Reply
It is a defining statement. Primary sources are generally firsthand accounts, or close to them. Secondary sources are generally secondhand ones. SlimVirgin TALK contribs 02:55, 11 March 2010 (UTC)Reply
Secondary sources are generally secondhand ones where we are talking about the facts from the primary source. But this is not the point of using secondary sources. Also, second hand accounts are not necessarily secondary sources. When Albert says that Barry said that Charles hurt himself, Albert's second hand testimony is not a "secondary source". --SmokeyJoe (talk) 03:00, 11 March 2010 (UTC)Reply
I'm not quite following what you're saying. A secondary source is a secondhand account. Easy example: I witness a traffic accident and I send you an email about it. That's a primary source. You take my email and you write an article about the accident for The New York Times. That article's a secondary source of information about the accident, because you weren't there. How to decide whether something's primary or secondary can get more complicated than that, but the above is the jist of it. SlimVirgin TALK contribs 03:08, 11 March 2010 (UTC)Reply
If I write a story about the accident, it's a secondary source, whether it is based on the facts from your email, or on the facts from my own earlier notes. It is secondary because I have added something to the facts. A secondary source is a secondary source due to the transformation of the primary source content into something beyond the primary source content. It is not a secondary source merely due to being second hand.
If I took your email and retold it in my own words, pretending that I was the eye witness, I have not created a secondary source; I have only created an unreliable copy of a primary source.
Alternatively, if you don't want to take my word, go to the text of the reference to the sentence in question. I think it clearly is closer to my copy edit than to what what you reverted to. Do you disagree? --SmokeyJoe (talk) 03:21, 11 March 2010 (UTC)Reply
SJ, that text about secondary sources has been there for a long time. It is correct. I think you need to leave it be and gain more experience of editing so you see how the policies work in practice. That will help you to see how these definitions translate in actual editing situations.
And yes, if you took my email, any article you wrote based on it would be a secondary source. SlimVirgin TALK contribs 03:28, 11 March 2010 (UTC)Reply
In the wider, non-wikipedian world, you are definitely wrong. You seem to hold the not uncommon, but mistaken view that any source that is not primary must be secondary. Now, you could (1) well defend the statement in wikipedia policy as something that describes what many longstanding wikipedians think. Or are you (2) interested in the real world usage of "secondary source"? Or are you interested in how the referenced statement is not a reflection of its reference? Or are you (3) interested in how it is detrimental to the encylopedia that many editors think that it is a good thing to build content on the basis of a reporter's non-transformative rendition of the facts. "It has been there for a long time" is not evidence of correctness. Instead, it is evidence of lack of review. --SmokeyJoe (talk) 03:41, 11 March 2010 (UTC)Reply
Perhaps it would help if you explain where your knowledge of sourcing comes from. If you have a PhD in history, for example, and you're telling me we've misunderstood what a secondary source is, that's something I'm going to take very seriously.
As for your example above, if Albert says Barry said Charles hurt himself, why should I not regard Albert's testimony as a secondary source, in your view? SlimVirgin TALK contribs 03:46, 11 March 2010 (UTC)Reply
SmokeyJoe, the Wikipedia usage of "primary source" and "secondary source" is completely different than in many real-life fields. For example, newspaper articles that report on an event just after it occurs would typically be considered primary sources (e.g. [1]) but here they are considered secondary sources. This is because we have declared by fiat that our articles are written using secondary sources, and thus we change the definition of "secondary source" to make this true. A huge amount of discussion has taken place about the PSTS section of this policy, and it's unlikely that any improvement is going to occur. — Carl (CBM · talk) 03:49, 11 March 2010 (UTC)Reply
Carl, you've taken that link out of context. It is specifically talking about the past. We also regard newspapers as primary sources in that context. This has been explained to you many times, but you're engaged in IDIDN'THEARTHAT. Again, Carl, if you have a PhD in history or similar, and are speaking from a position of authority, please let us know. At some point we need to trust people who have actually studied this issue. SlimVirgin TALK contribs 03:52, 11 March 2010 (UTC)Reply
SlimVirgin, I have explained this to you several times as well, but you continue to ignore it (throwing rocks with glass walls...)
Articles about events yesterday are indeed articles about the past; there is no special exception that says that articles about the Vietnam war written in the 1960s are primary, but articles about the Iraq war written in 2010 are secondary simply by virtue of being more recent. I am not inventing a new definition of "secondary source" here; I am simply pointing out the way the term is usually used. Wikipedia uses it in a different way, which has the effect of simply redefining the sources that are favored here as "secondary" and redefining sources that are disfavored "primary". — Carl (CBM · talk) 03:57, 11 March 2010 (UTC)Reply
Several editors with Humanities degrees, including postgraduate degrees and degrees in history, have explained to you that what you're saying is wrong, including people who otherwise disagree with each other. You have never been able to produce a source supporting your definition. SlimVirgin TALK contribs 04:01, 11 March 2010 (UTC)Reply
You should read the link I included above: [2]; I have shown you similar sources before. I have looked at many sources during the course of our discussions and I have a good sense for the nuances in them about classifying primary and secondary sources. It's perfectly standard, although not universal, to classify news reporting that is published during the time period of an event as a primary source about that event. We don't follow that here, however.
Now, it isn't true that "Several editors with Humanities degrees" have explained anything to me, at least not while pointing out to me that they have humanities degrees. The only person I recall explaining your side of the story about secondary sources to me is you, and frankly I believe you are quite biased regarding several of the content policies.
On the other hand, no editor with a humanities degree could read the sentence "... this includes unpublished facts, arguments, speculation, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position." with a straight face. — Carl (CBM · talk) 04:17, 11 March 2010 (UTC)Reply
(ec)I was a scientist who has moved into one of the soft humanities. I do not have a PhD in history, nor am I a historian, but have had a long interest in both history and historiography. I don't want to talk about myself, but tell you this because it gives me a very clear perspective on how imprecisely the humanities defined their terms. I learned the meaning of "secondary source" from others and from using it studying history. I'll now ask you to believe that in my opinion the best reference for "secondary source" is the Wikipedia article Secondary source, including its references. What I have trouble with is with how many people fail to see the inconsistency between the lead of that article and the section in this policy.
To quote from Secondary source. "a secondary source ... relates or discusses information originally presented elsewhere". If Albert's testimony doesn't relate or discuss, if it merely repeats, then it is not a secondary source. If Albert's testimony is a verbatim copy of of Barry's testimony, then it may as well be a photocopy of the same piece of paper. To the extent that Albert's testimony differs from Barry's in terms of the details of Charles' injuries, the differences are probably well described as unreliable. I am talking about Chinese whispers here. If I may digress to my preferred angle (#3 above), encouraging the building of content on the recording of the results of Chinese whispers is bad thing. --SmokeyJoe (talk) 04:18, 11 March 2010 (UTC)Reply
And once again, we get into a debate about how to define Primary vs. Secondary (interesting that we never argue about Tertiary)... Slim, there is obviously something wrong with our definition because this keeps coming up. It isn't just clueless newbies. I don't know exactly what is wrong, but the simple fact that this section keeps getting questioned tells me we have a problem. Blueboar (talk) 04:41, 11 March 2010 (UTC)Reply
You were one of the people correcting Carl before, BB. It's always the same people and they congregate here because they don't understand it. I see the definitions being used well on WP every day. SlimVirgin TALK contribs 05:09, 11 March 2010 (UTC)Reply
Slim, before accusing people of "congregating here" you should look at how often you edit the NOR policy yourself. I understand both the policy and the actual references on primary and secondary sources perfectly well; saying that they are at times in conflict is not a sign of misunderstanding anything or of not hearing explanations.
The deeper issue with PSTS (and I think Blueboar has agreed about this in the past, although he is free to correct me) is that it has little to do with "original research". If material is accurately sourced, not going beyond what is in the sources provided in any way, then it is not original research, even if it might violate WP:BLP, WP:NPOV, and generally be a bad idea. This holds regardless whether the source is primary, secondary, etc. — Carl (CBM · talk) 05:24, 11 March 2010 (UTC)Reply

Hoping to make one small step forward. Looking at the following:

'''[[Secondary sources]]''' are second-hand accounts, at least one step removed from an event. They rely for their material on primary sources, often making analytic or evaluative claims about them.<ref>[http://www.lib.berkeley.edu/instruct/guides/primarysources.html University of California, Berkeley library] defines "secondary source" as "a work that interprets or analyzes an historical event or phenomenon. It is generally at least one step removed from the event".</ref>

There is self-inconsistency in this. The reference does not say "Secondary sources are second-hand accounts", but that they are "generally", "one step removed". This logically means that that there can be secondary sources that are not one step removed. Also, it does not say "second-hand accounts are secondary sources" which is what some people walk away with having got to the end of the sixth word.

But I am not here just to be pedantic. That part of the policy encourages the building of content based on "second-hand accounts" and this is a bad thing. Often, the second hand account is a newspaper report, and it is devoid of commentary, analysis, or any other transformation of the facts. Such a source is not a source for scholarly reading. It is just a compilation of primary source material. It is not good foundation for an encyclopedia article. Reporters report almost indiscriminately, depending on other news of the day, and where the reported happened to be.

The other failing of policy in this is the use of "reliable". "Reliability" is not a quality of secondary sources. Primary sources are reliable or unreliable. The facts they report are accurate or not. Secondary source material is not right or wrong. Secondary source material is the author's opinion (perhaps their analysis is wrong, but here we get into usage). The adjective for a good quality secondary source is "reputable". --SmokeyJoe (talk) 05:53, 11 March 2010 (UTC)Reply

SJ, you've made 251 edits to articles in four years of editing. With the best will in the world, it isn't advisable to make substantive edits to a core content policy without experience of how the policies function in practice. Something that might make no sense to you at present might suddenly start to make sense after a year of solid article editing. SlimVirgin TALK contribs 05:57, 11 March 2010 (UTC)Reply
This response is addressed to SmokeyJoe. To be fair, the use of newspapers is much less common in a<script type="text/javascript" src="http://en.wikipedia.org/w/index.php?title=User:Omegatron/monobook.js/addlink.js&action=raw&ctype=text/javascript&dontcountme=s"></script>reas of wikipedia where there are scholarly sources available. For example, I work mostly in mathematics, and we almost exclusively use textbooks and journal articles. The same is true in the natural sciences. For articles in these areas, there is usually no difficulty separating reputable sources from nonreputable ones, and there is rarely any reason to use a newspaper article as a source.
The unique challenge that Wikipedia has is trying to write sourced articles about current events for which scholarly resources are extremely limited, such as current events and people in the news. The compromise that was made is to allow the use of newspapers as the fundamental sources for these articles. This compromise is necessary, in my opinion, because there simply aren't other sources available. The downside is that these articles too often devolve into a series of staccato sentences, each sourced to a different newspaper article or website.
In any case, the fact that many articles are written from newspapers is here to stay. The NOR policy handles this situation by defining these sources as "secondary" so that we can say that all articles should be written using secondary sources. This does lead to the unfortunate result that "secondary sources" in NOR are not the same as what many people would call "secondary sources" in their professional work. However, once you realize that this is what's happening, and the motivation behind it, it's not indefensible, even though it is unfortunate and confusing to new editors. — Carl (CBM · talk) 06:10, 11 March 2010 (UTC)Reply
Your understanding of primary sources is just wrong, Carl, and given that your focus is on mathematics, you might consider showing just the minimum of respect to people who have studied sourcing formally. If I were to start contradicting you about mathematics—and kept on doing it year after year after year, so that whenever you posted X, I'd pop up out of nowhere to post not-X, even though I had no training in mathematics and could produce no sources in support of my position—you might wonder what I was up to. :) SlimVirgin TALK contribs 06:23, 11 March 2010 (UTC)Reply
I'll try bold: HERE IS A SOURCE THAT SUPPORTS MY POSITION: [3]. I have spent some time looking at sources like this as part of these discussions, and I actually do look at the sources that other people give me, so I have a pretty reasonable sense of what is in the references about primary and secondary sources. In particular, I am aware that both "newspaper articles are primary" and "newspaper articles are secondary" can be found in reputable sources. So there is no need to try to change my mind about that.
Also, I am not "popping out of nowhere"; I have this page on my watchlist and have had it for some time. I keep responding to SmokeyJoe, and you keep deciding that you want to argue with what I say. That's your prerogative, but there's no requirement that I need to agree with you in order to explain the NOR policy to other people. Perhaps you should simply let me discuss with Smokey and find something more productive to do than argue with me. It's not as if I am editing the policy page; I'm simply explaining to a new user the motivations behind it – motivations that you have emphasized are important for anyone who wants to edit the NOR page. — Carl (CBM · talk) 06:35, 11 March 2010 (UTC)Reply
For the umpteenth time, Carl, that source does not support your position. It is only discussing sources from the past, and we agree with it. This has been explained to you many times by many editors, so please stop making us say the same thing over and over. Perhaps you could write to that source and ask for clarification about how they would classify a newspaper article by an uninvolved journalist about an event that happened yesterday. You might believe it if it comes from an uninvolved academic. :) SlimVirgin TALK contribs 07:36, 11 March 2010 (UTC)Reply
It may be that you don't understand my position. My position is that there are plenty of references that classify newspaper articles produced contemporaneously with an event as primary sources about the event. The source I provided certainly supports that. There is no magical distinction in th source there between events that happened yesterday and events that happened 20 years ago: they are both historical events from the point of view of someone doing historical research.
Here are two more refs, as long as you are claiming they don't exist:
  • Berkeley: "Primary sources were either created during the time period being studied or were created at a later date by a participant in the events being studied (as in the case of memoirs). "
  • Western Carolina U: "Newspaper articles can be either a primary or a secondary source. For example, when a newspaper article initially reports on an event, the newspaper serves as a primary source. When a newspaper article reflects back on an event, it is a secondary source."
  • James Cook University "A newspaper article is a primary source if it reports events, but a secondary source if it analyses and comments on those events."
  • Harvard public library "A magazine or newspaper article published during the time of the topic would also be a primary source. For example if the topic is about the space shuttle Challenger explosion a newspaper article from January 29, 1986 would be a primary source."
Of course there are also sources that any newspaper articles are secondary sources in general, but it's unconvincing when you tell me that the things I am reading in actual references don't exist.
I think the real difference in our positions about this is that you are claiming that there is some distinction between "recent history" and "non-recent history", but that distinction will not stand up upon closer inspection. In any case, that distinction is not really related to the NOR article on Wikipedia; from our point of view, writing an article on an event from last month is not really different than writing an article on a news event from 1960.
However, like I said, it may be easier if you simply avoid responding to posts that were not aimed at you in the first place. If, as you claim, you are tired to telling me things, then simply stop telling them to me. — Carl (CBM · talk) 12:41, 11 March 2010 (UTC)Reply

arb break

 
SV, why do you want to discuss the person, not the issue? I am not here editing policy, but contributing to a talk page discussion on a focused issue. I am discussing something that I understand very well. But if you want to talk about people, I think this discussion and this graphic suggest that this page is being WP:OWNed by someone. --SmokeyJoe (talk) 06:21, 11 March 2010 (UTC)Reply


Carl, I understand all that, and I am not here to try to prevent the use of newspapers. I understand that none of this is relevant to obviously scholarly subjects. In scholarly areas, we don’t refer to policy on how to write content. It is relevant to dubious subjects only covered twice by a local newspaper. It is relevant to many articles that appear at AfD where newcomers are misled by the sometimes acceptance of online news reports as the basis of an article.

Correcting the illogic of this policy need not change any practice. We can easily be explicit that the appearance of a story in two regional newspapers demonstrates notability.

I’m glad you recognise the confusion to new editors. This is an issue of accessibility of the project to newcomers. SV seems to suggest that I cannot understand this core policy without a solid year of mainspace editing. I think that is not good. --SmokeyJoe (talk) 06:31, 11 March 2010 (UTC)Reply

I didn't say that. I said it's difficult to change it without editing experience and knowledge of how it works with the other policies. SlimVirgin TALK contribs 06:38, 11 March 2010 (UTC)Reply
It's difficult to change if you are going to keep it soft protected and then talk about my edit history when I try to talk about the suggested change. --SmokeyJoe (talk) 10:43, 11 March 2010 (UTC)Reply
Re SmokeyJoe: OK, great. I wanted to point out the first stuff first. I have to admit I am very pessimistic about the possibility of really clarifying the content policies, because there is a lot of institutional inertia behind them. But I can explain why they are written the way they are, at least.
Do you think it would be clearer if the policy said something like "the classification of sources as primary or secondary here is specific to wikipedia, and may differ from those an editor is familiar with"? For some reason, many of our policies re-use existing terms in idiosyncratic ways. — Carl (CBM · talk) 06:45, 11 March 2010 (UTC)Reply
I think we need to avoid, or minimise, wikipedia-specific terms. --SmokeyJoe (talk) 07:24, 11 March 2010 (UTC)Reply

<outdent..edit conflict etcetera>I reject the idea that re-examining the def's of secondary/primary sources is worth the effort for the following 2 reasons: 1) been there, done that, even if this one isn't perfect, more perfect options aren't demonstrably any better, and 2) I do not see that this "confusion" lies at the heart of most NOR disputes.

I believe novel interpretation, synthesis and undue weight conflicts rebound to the "secondary source" question - nobody cares otherwise if a source is "primary" or "secondary". It is not an infrequent problem that wikipedia editors are here pushing to establish a "definitive" take on a given issue. And ignoring the broadstream/mainstream views to get to the quote/unquote "truth" of things, blowing past the secondary source material, is a frequently used end-run around disputes over "which are the most reliable sources" for a particular claim given in articles here on wikipedia or "which sources are biased and which are reliable". WP has made up its mind--it is a free encyclopedia, not a pioneering champion of "DIY truths". No amount of tinkering with the primary/secondary source question will smooth these wrinkles...because they're not really a significant wellspring of dispute. Primary/secondary source "confusions" are NPOV and RS disputes, disguised. Professor marginalia (talk) 07:48, 11 March 2010 (UTC)Reply

1) We've been here before? Yes, we have. There's been little progress. Wikipedia:Primary, secondary and tertiary sources was a serious attempt to more forward, but today, SlimVirgin is attempting to shut it down and revert here to old text.
2)This confusion does not lie at the heart of WP:NOR disputes, because the heart of the issue was sectioned off to WP:N.
Professor marginalia makes a number of statements about how solving this policy redefinition issue won't change much. He's possibly right, but having self-inconsistent policy certainly doesn't help. --SmokeyJoe (talk) 10:05, 11 March 2010 (UTC)Reply

IMO the one thing that is wrong with the primary/secondary/tertiary source distinction is that we are making it at all. Let's take a newspaper report about a notable scientist dying in a traffic accident, for example. If the accident happened last week, we must use such sources. However, if the notable scientist was Pierre Curie it's a lot better to use one of his biographies. Either both newspaper reports are primary sources, or both are secondary sources, or we have a weird definition that makes a secondary source mutate into a primary one in the course of 100 years.

I recognise that the distinction is an attempt to make explicit some of the criteria that a competent editor uses implicitly when comparing the value of different sources, so that editors can communicate more clearly and consensus can be found faster. I don't think it is making a good job of it: It is too complicated and too many are using it as a rhetorical weapon rather than an analytical tool.

To continue the Pierre Curie example, while making it more hypothetical: Suppose all biographies agree that he was killed by a horse-drawn carriage, but L'Aurore and La Gazette reported it was an 'omnibus'. The (real or perceived) discrepancy has led to a battle between two entrenched factions. One side insists that newspaper articles are secondary sources, assume that the nature of a source can't change, and insist that since the newspapers have much more detail than the biographies (tertiary sources?) they are the better sources. The other side insists that old newspaper articles are primary sources and therefore we must build our article on the biographies, which are secondary sources. There is never a definite answer, and the discussion continues to revolve around this technicality rather than asking questions such as whether 'omnibuses' in Paris in 1906 were horse-drawn carriages and which version is more useful and more accurate for a modern audience.

In this hypothetical conflict everybody learns one thing: Nobody really knows what a primary or secondary source is. You just pick which sources you want to use, and then you argue that they are secondary and any contradicting sources are primary or tertiary. I believe this is how it happens in POV battlefields, and it's not a surprise because as I have explained this is what most editors learn in the less contentious areas, where no Wikipedia rules expert is around who could explain how to apply the rules in a specific case so that they make sense. (And often even if such an editor is around, they will focus on resolving the conflict rather than teaching a general lesson.)

Perhaps all this primary/secondary/tertiary stuff has fulfilled its purpose now by making us all aware of certain aspects that need considering, and should simply be scrapped now in favour of something radically simple such as picking the sources that are most likely to be right for the specific application. I really think it would reduce the wikilawyering. Hans Adler 07:50, 11 March 2010 (UTC)Reply

PS: I was talking about the general Wikipedia audience. Of course the editors of an article decide what is the most efficient way to communicate about the relative quality of sources, and at certain humanities articles the primary/secondary/tertiary distinction may be an important part of that because everybody is familiar with it from work/study. But not at Pokémon, Beer, Evolution, Leeds, Hankel singular value or Chicago Tribune Silver Football. Hans Adler 07:57, 11 March 2010 (UTC)Reply
Taking your example of Pierre Curie, it is not better to use his biographies or contemporaneous newspaper articles. It's better to use secondarily sourced materials published by historians exercising disciplined distance, whose veracity is most widely accepted. WP doesn't do research...it prepares a free encyclopedic definition which is derived by published researchers. Professor marginalia (talk) 08:10, 11 March 2010 (UTC)Reply

I am a little dismayed by the number of credentialism-oriented remarks in this section. I'd have thought that the most experienced editors would have a long enough memory to not make that mistake again. We don't care who the editor is, what real-world claims he makes about his identity, or how much he (or she) has done (under this username): we care whether a given change makes this page more perfectly represent the community's current view.

And, yes: This is a hard section to get right. It's hard because the scholarly definition of 'secondary source' changes when you move from history to science to popular culture; it's hard because Wikipedia's definition does not entirely match any of these real-world definitions; it's hard because we have to address such a wide range of skill levels and education among our editors; it's hard because the community uses 'primary' and 'secondary' as code words for 'bad' and 'good' in ways that have nothing to do with their actual status as primary or secondary in the relevant discipline; it's hard because the community isn't entirely consistent about what it wants.

But I do not think that whether or not a person has a PhD in history, or a zillion mainspace edits, or any other set of credentials determines whether or not a change accurately describes the community's views. Every change should be evaluated on the merits of the change, not the assertions of the editor's identity or standing in the community. I hope that there will be far less commenting on editors's credentials or perceived ownership problems, and far more focus on content. WhatamIdoing (talk) 09:02, 11 March 2010 (UTC)Reply

I don't see why Wikipedia should be the only place in the world where credentials don't matter. I don't want to be told about sourcing by someone who's never studied it and doesn't understand it, just as I don't want to be told about mathematics by someone with a degree in history, or lectured on how to use sources on WP by someone who's made 251 edits to articles over four years. Someone with no experience of adding content can't know what the community's views are on the content policies, or how the policies work in practice. Similarly, no one would want me to try to teach them how to write computer code, or create a template, or close a deletion debate, because these are things I've never studied and have no experience of. No one insults me by pointing that out. It's just a fact.
I can tell you that there would be very strong wiki-wide opposition to weakening any aspect of the policy to encourage the use of primary sources, not because there's anything wrong with them, but because a lot of Wikipedians don't know how to use them, and misuse of primary sources is one of the more common kinds of OR. SlimVirgin TALK contribs 09:28, 11 March 2010 (UTC)Reply
Oh, and you mentioned Essjay. One of the very early signs that he didn't have the degrees he said he did was that he didn't know what a secondary source was—no one could get through a PhD in the Humanities without knowing that. Anyone who believed him really wasn't reading what he wrote (in fact it was so obvious that to this day I believe Essjay intended it as a joke, not as a real deception). SlimVirgin TALK contribs 10:01, 11 March 2010 (UTC)Reply
If you want to make some claim about credentials, you need to make it explicitly. However, as I said above, I would find it surprising that that someone with an advanced humanities degree would be able to stomach the claim "This includes unpublished facts, arguments, speculation, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position. " which defines all our writing as original research, because all writing advances a position. This is the second sentence of NOR right now. I have always thought it is interesting the this policy is so clearly not written from an academic perspective. I have always taken things like the sentence I quoted as evidence that there are not humanities experts applying their professional knowledge to this policy. — Carl (CBM · talk) 13:24, 11 March 2010 (UTC)Reply
You ought to read the sentence you're quoting—all of it, instead of just one part of it. I wonder how many years you'll keep this going, Carl. :) SlimVirgin TALK contribs 13:27, 11 March 2010 (UTC)Reply
The quote is an entire sentence from the policy: "This includes unpublished facts, arguments, speculation, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position." The fact that this defines all writing as original research makes me smile each time I read it. — Carl (CBM · talk) 13:30, 11 March 2010 (UTC)Reply
I know the quote is an entire sentence. I can see it. But you're not reading it properly. Anyway, Carl, for years, if I've said black, you've said white, and you must always have the last word so I'll leave it to you now. I think one day I'm going to have to arrange a little breaching experiment, and suddenly start agreeing with you, then watch your head explode as you try to backtrack. :) SlimVirgin TALK contribs 13:33, 11 March 2010 (UTC)Reply
Thanks, I like having the last word. Does that mean you will stop trying to correct me simply because you disagree with me?
The clause "any unpublished analysis or synthesis of published material that serves to advance a position" is perfectly clear. Because all writing advances a position, if Wikipedia articles count as unpublished (as I am sure the policy intends), this clause says they are original research. Anyone familiar with literary analysis, and any professional historian, would notice this immediately. You said Essjay's misunderstanding of secondary sources made it clear he was not an academic historian; the existence of that sentence in the NOR policy has always made similar things clear to me about policy editors in general, although not about any editors in particular, because I never looked up who wrote the sentence originally. However, because the policy does not even get basic things like that right, it makes sense that it might also not use terms like "secondary source" in the way they would be used by academics. — Carl (CBM · talk) 13:41, 11 March 2010 (UTC)Reply
I'm sorry, I can't resist. Your analysis of that sentence is so wrong-headed I don't know where to begin, but I will say this: most of this policy was written by people with post-graduate degrees in the Humanities, including PhDs. The writing is not the way it ought to be because others arrive to tweak it around and then it becomes impossible to change. But the ideas are solid and they work. This policy together with V and NPOV can solve just about any editing dispute—no matter how complex—when applied correctly, and that's what matters. SlimVirgin TALK contribs 13:52, 11 March 2010 (UTC)Reply
I think the problem is that while SlimVirgin correctly sees that the policy can be read and applied correctly, she doesn't see how easy it is for it to be misunderstood. I think it has been forgotten who is the intended audience for the policy. It certainly shouldn't be written for the peers of the writers, but for the newcomers. Advice for the writer: It is not good enough that the text is correct and understandable. It must be nearly impossible to misunderstand or misinterpret. SlimVirgin, I think you are not taking on board the feedback that the text is too easily misunderstood and misinterpreted. --SmokeyJoe (talk) 01:55, 12 March 2010 (UTC)Reply
Yes, that's the issue. Policies like this need to be written with a level of clarity that allows even complete newbies and determined POV pushers to figure out that their desires don't conform with the policy. We need a page that serves our primary purpose (=educating editors that don't know the answer already). WhatamIdoing (talk) 23:46, 12 March 2010 (UTC)Reply
I think the problem is that you are reading NOR using Wikipedia-words instead of real-world-words. In real world words, whenever I write an article I am synthesizing material from the sources to create a new document, and that document will always advance some position (because all writing advances a position). Anyone taking history 101 would learn these things; they're just a fact of life independent of any Wikipedia policy. However, when people read the NOR policy here, its plain language says that synthesizing written material to advance a position is classified as "original research".
This is just another example where the policy wants to say something, but doesn't achieve it. People who know how to substitute Wikipedia-meanings for the words of the policy can make sense of it, but for people who try to read the plain language without viewing it as a form of coded speech often find it confusing of self-contradictory. Therefore, the main task in explaining policy to new editors is teaching them how to recode the words in the policy into their Wikipedia-meanings.
Regarding credentials, like I said, you can either claim to have them them or not, as you like, but vague claims of "post-graduate degree in humanities" are never going to be convincing. Particularly when there are such basic errors in the policy text; like I said, any professional academic historian would instantly recognize the problem with the sentence I pointed out. — Carl (CBM · talk) 14:09, 11 March 2010 (UTC)Reply
Read the sentence you're analyzing! Articles may not contain unpublished facts, arguments etc, or and any unpublished analysis or synthesis of published material that serves to advance a position. Any such material must have been published before. You've made 355 posts to this page since 2006, almost all of them revolving around the same two misunderstandings, and no matter how often they've been explained to you by multiple editors, you just keep repeating them. :) SlimVirgin TALK contribs 14:21, 11 March 2010 (UTC)Reply
The article itself is an unpublished synthesis of material that serves to advance a position. (On the other hand, if the article is "published", then the synthesis in it is also published...) The entire article – every word of it – is synthesis that advances a position. This is because all writing is that way.
The NOR policy relies on Wikipedia-meanings to make sense. What the policy actually wants to say is that the arguments made by the article need to agree sufficiently, in the eyes of the editors of the article, with the arguments made by the sources of the article. But even when these are in perfect agreement, the article will still consist of unpublished synthesis of its sources. So the things I am pointing are not "misconceptions", they are rather idiosyncracies in the way that this policy is written that keep it out of line with the actual meanings of the words it uses.
Your personal attacks are somewhat tiring: I could equally well claim that no matter how many times things are explained to you, SlimVirgin, you ignore them and the sources provided to back them up :). Claims like that accomplish nothing. Personally, I understand the policies and their motivations quite well, and I am very experienced with how sourcing actually works on Wikipedia. Discussing how to get the policy to say what it means while using terminology in a way that is recognizable to new readers is in no way inappropriate. — Carl (CBM · talk) 14:39, 11 March 2010 (UTC)Reply
The aim is to write articles that don't advance positions that haven't been advanced by others. And WP articles are not unpublished. All the things you're raising, and have been raising since 2006, are Editing 101. That's not a personal attack. I just don't see the point. SlimVirgin TALK contribs 14:46, 11 March 2010 (UTC)Reply
Perhaps you are reading "unpublished material" to mean "previously unpublished" material, but that isn't what the NOR policy actually says. If the contents of our articles are "published" then our articles do not contain unpublished material; that's just a tautology. However, even if you replace "unpublished" with "previously unpublished" (which is what NOR is trying to say), it's still true that the content of our articles is "previously unpublished material that advances a position", unless the articles are copyvios. That is: (1) the material in our articles is previously unpublished and (2) the material in our articles advances a position. Which of those two do you disagree with? — Carl (CBM · talk) 15:44, 11 March 2010 (UTC)Reply
Of course I agree that "The aim is to write articles that don't advance positions that haven't been advanced by others." My point is just that that is not what the second sentence of the NOR policy says. It says our articles cannot advance a position at all, which is impossible. I would be happy to see "The aim is to write articles that don't advance positions that haven't been advanced by others. " inserted into the NOR policy, since it is much more clear. — Carl (CBM · talk) 15:47, 11 March 2010 (UTC)Reply
But that's exactly what that sentence says. I think I'm going to start collecting all the times you've argued black is white, and supported any editor, even if they registered five minutes ago, so long as they disagreed with me. :) In the course of looking around, I found the last time several people explained the "newspapers can change from secondary to primary depending on the distance from the event" issue: see here. Blueboar explained it particularly clearly in the last post in this section, and you didn't respond, so I'd hoped that was that. How wrong I was! :) SlimVirgin TALK contribs 16:00, 11 March 2010 (UTC)Reply
The sentence in NOR does not say, "The aim is to write articles that don't advance positions that haven't been advanced by others." It says, "any unpublished analysis or synthesis of published material that serves to advance a position." On the one hand, our articles are published, so they cannot include unpublished material by definition. However, if we read it as "any previously unpublished analysis or synthesis of published material that serves to advance a position." then every article violates this, because every piece of written text serves to advance a position. This is not a difficult thing to understand, so my impression is that you simply are not reading the NOR language in its plain sense. Are you actually disagreeing that every piece of writing advances a position? That would be a bizarre argument for anyone familiar with historical or literary analysis. — Carl (CBM · talk) 16:12, 11 March 2010 (UTC)Reply
You just keep repeating yourself as though nothing has been explained. You don't incorporate the explanations, and say, "yes, but," or "ah, but I disagree because ..." You just post almost word for word what you've been posting for years. : ) You surely know that the sentence means no previously unpublished analysis/synthesis, but as I've said elsewhere we can't fix the writing because of editors like you. Perhaps you don't mean to be having that kind of effect but it's what's happening, at least from my perspective. I know that you will object to any rewrite, and so all I dare do is tinker around trying to stop the writing from deteriorating any further. SlimVirgin TALK contribs 16:33, 11 March 2010 (UTC)Reply
I am not a new editor looking for an "explanation". I understand the actual policy very well, and I understand what is written on the policy page, and I am simply pointing out they do not agree. You have not explained in any way your disagreement with "every piece of writing advances a position". It is that issue that I am talking about here; the "previously published" is a more minor thing that you could certainly fix with no objection from me. I am not asking you "what does the sentence mean" – I am telling you that the sentence does not say what we want it to mean. — Carl (CBM · talk) 16:43, 11 March 2010 (UTC)Reply
Also, regarding the PSTS section, I have been actively agreeing with proposed rewrites for years, as have other editors here. The new separate page for PSTS seems like a reasonable move – and I was uninvolved with that. Rewriting the first paragraph to say what it is supposed to mean would be a great improvement in the policy page. — Carl (CBM · talk) 16:45, 11 March 2010 (UTC)Reply
I fail to understand your point, CBM. The current wording agrees exactly with actual policy, and the sentence says exactly what we want it to mean. Jayjg (talk) 01:48, 12 March 2010 (UTC)Reply
Let me say the point again: because every piece of written text advances a position, and the sentence in NOR says that any article that advances a position is original research, the sentence in NOR says that every article is original research. — Carl (CBM · talk) 02:00, 12 March 2010 (UTC)Reply
Yes, I understand that you've said that, but what does it mean? If someone writes in an article "Billy Crystal was born in Doctor's Hospital in Manhattan", what "point" does it "advance"? Jayjg (talk) 22:37, 12 March 2010 (UTC)Reply

SV, I do not see why a statistic of this user account is relevant to answering a challenge to an inconsistency in the text of this policy.

The inconsistency can be stated thus: The WP:PSTS definition of “Secondary source” is inconsistent with the mainspace article it boldly links, and the formal references to sentences two and three.

I am not lecturing you on how to use sources. You seem to use them fine. The problem is with the text you reverted to as part of a large revert, which you only attempt to justify by bald assertion, and “has been there for a long time”.

The community's views are on the content policies, and how the policies work (and don’t work) in practice are as plain as day to see. Why you bring this up, I don’t know. No one here is trying to weaken policy.

SV, you don’t know me, you don’t know my credentials (if any), and you don’t know what I know. So why don’t you forget about who’s saying it, and look again at that easily misread implication in the text that “second-hand accounts are secondary sources”. Surely it can be written more clearly. Failing all else, why not quote the reference explicitly? --SmokeyJoe (talk) 10:45, 11 March 2010 (UTC)Reply

Clutching at straws here trying to understand SV's hard reaction... I have no connection, past or present, with User:Essjay. --SmokeyJoe (talk) 10:49, 11 March 2010 (UTC)Reply
I didn't say you had anything to do with Essjay; please stop putting words in my mouth. What I'm saying is that either (a) you really do only have 250 edits to articles in four years, in which case you can't know how to write or apply the content policies, or (b) you're a sockpuppet/alternative account, and it's a violation of SOCK to edit policies with a second account. Either way this is time-consuming and unhelpful. SlimVirgin TALK contribs 10:58, 11 March 2010 (UTC)Reply
I dare say I have a strong track record in working towards the application of content and other policies at MFD, AFD and DRV. Rarely do I find myself standing alone. I see the test cases often, the arguments, and the bitten and confused newcomers. I'm telling you that of all the things on this policy page, the wording associated with mis-re-defining of "secondary source" is a cause of needless problems and could be easily fixed.
I assure you that I have have never violated the spirit or the words of WP:SOCK.
I reject the notion that attempting to improve the comprehensibility of core policy is a waste of time. I have no doubt that you understand what you think it says, but I don't think you understand how others read it differently to you. Normally, you do very well in writing policy, and I agree with you that style and flow of important documentation is not helped by many small edits, but here there is a small but definite problem. Also, looking at the discussion, and this archives and subpages of this page, I do not feel that I am alone. You seem to resent having a piece of your creation criticised. I understand that, and am sorry, but a second hand account does not make a secondary source. --SmokeyJoe (talk) 11:36, 11 March 2010 (UTC)Reply
My recollection is that I didn't write that section originally, though I could be wrong because it was so long ago. But okay, I've said my piece, so I'll leave it to others now. SlimVirgin TALK contribs 11:44, 11 March 2010 (UTC)Reply
SV says: "...there would be very strong wiki-wide opposition to weakening any aspect of the policy to encourage the use of primary sources, not because there's anything wrong with them, but because a lot of Wikipedians don't know how to use them, and misuse of primary sources is one of the more common kinds of OR." (Italics mine)
That is the problem in a nutshell... There isn't anything wrong with using primary sources... but there is something wrong with misusing them. We don't really address that issue in the policy. We spend so much time and effort explaining to the reader how to determine whether a source is primary, that we never get around to explaining to the reader how to determine whether a misuse is occuring. Blueboar (talk) 13:13, 11 March 2010 (UTC)Reply
We can't do that for the same reason that we can't make the writing in the policy clearer: the people who tend to congregate around this page object to everything. So the current state is the best one that's possible in this climate. SlimVirgin TALK contribs 13:20, 11 March 2010 (UTC)Reply
What is that supposed to mean, Slim? You’re objecting to changing the policy because we can’t change it because people keep objecting to changes. Eh?
In terms of dealing with primary, secondary and tertiary sources, I think this is much better handled by the proposed guideline WP:Primary, secondary and tertiary sources. It doesn't actually contradict anything that is said in WP:NOR but it has a different emphasis, which I think solves a lot of the problems we are talking about.
And hey, if it doesn’t make things clearer then at least next time the big argument about it will be on its talk page, rather than clogging up WT:NOR!  :-)
Yaris678 (talk) 14:57, 11 March 2010 (UTC)Reply
The proposed guideline was yet another attempt to loosen up the NOR policy, in this case to allow people to make liberal use of primary sources. One of its main objectives was to actually mandate the use of primary sources in certain cases. It is a failed guideline, and for good reason. Jayjg (talk) 01:48, 12 March 2010 (UTC)Reply
Just to clarify, since I have apparently not been entirely clear enough to convince people that I mean this: I really do mean that I do not care what the editors' credentials are. If you take a net-connected laptop to your local zoo, and a Cebus capucinus mashes the keyboard and randomly happens to improve this page, then I'm not going to oppose the improvement simply because the monkey is incapable of reading English. WhatamIdoing (talk) 21:48, 11 March 2010 (UTC)Reply
Well, I do care when inexperienced editors try to change policy, because they invariably don't understand it well, and make a hash of it. Jayjg (talk) 01:48, 12 March 2010 (UTC)Reply
I care (very much) about hash-making, but I do not agree that inexperienced editors invariably fail to understand the policy or invariably make a hash of it. (I believe that it's likely, but not the invariable outcome.)
Similarly, I've seen highly experienced editors make a hash out of Wikipedia's guidelines and policies. An edit count is not a good determinant of whether a person is a good policy writer. You can deeply understand the policy without being capable of communicating your understanding to people that don't already know the answer. WhatamIdoing (talk) 23:42, 12 March 2010 (UTC)Reply
It's not an issue of edit count, but of experience with content. The content policies must serve the needs of content contributors, and you can't know what those needs are unless you're one of them. Editors come here to show that their edits are policy compliant, or that other edits aren't. They need to know that the words and phrases they need are here. An editor with no experience of contributing content can't know what those words and phrases are. So what might appear to be a reasonable copy edit could be disastrous in terms of the function of the policy. SlimVirgin TALK contribs 23:54, 12 March 2010 (UTC)Reply
We're not mind readers. We often can't know what someone else does or does not know.
I care little about whether a source is primary or fourthary. :) But I do care that discussions should be centered on the issues and not the people involved. Maurreen (talk) 00:21, 13 March 2010 (UTC)Reply
If someone arrives to change the policy and has made only 250 edits to articles over four years, we do know that either (a) that person is not a content contributor or (b) is a sock/alternative account, and policy editing with second accounts isn't allowed. So either way it's unhelpful and time-consuming. That's demonstratively true—look at how many words have been spent on this, and do any of you still remember the proposed edits? It's fine for those who don't care who is doing what, but others do. I do care because I don't want to see key words and phrases removed, and the names of the editors helps me to judge whether the policy is being reformulated by someone who knows their stuff. That's all I want to say about it. This really isn't a fruitful discussion. SlimVirgin TALK contribs 00:27, 13 March 2010 (UTC)Reply
While the context / metadata of who the editor is of some relevance, I think that being dismissive of a contribution or implying a sock account for only having 250 edits possibly takes that too far. The amount of time a person spends on Wikipedia is not necessarily a measure of (or even correlated with) the quality of a contribution. Respectfully. North8000 (talk) 02:08, 13 March 2010 (UTC)Reply

newspaper articles are primary/secondary

Above someone said: "both "newspaper articles are primary" and "newspaper articles are secondary" can be found in reputable sources." Can someone point me reputable sources that say either of these things? --SmokeyJoe (talk) 10:47, 12 March 2010 (UTC)Reply

This book names obituaries and 'histories written at the time' as examples of primary sources. Both of those are specific types of articles that we'd expect to find in a modern newspaper. (Have you read Primary source?) WhatamIdoing (talk) 00:06, 13 March 2010 (UTC)Reply

I tried to tidy some writing and was reverted by SmokyJoe who said I had made it "erroneous". Perhaps he could say more.

Current version My copy edit
Reliably published tertiary sources can be helpful in providing broad summaries of topics that involve many primary and secondary sources. Some tertiary sources may be more reliable than others, and within any given tertiary source, some articles may be more reliable than others. Wikipedia articles may not be used as tertiary sources in other Wikipedia articles, but are sometimes used as primary sources in articles about Wikipedia itself. High-quality tertiary sources can be helpful in providing broad summaries of topics. Wikipedia may not be used as a tertiary source, but is sometimes used as a primary source in articles about Wikipedia itself.

SlimVirgin TALK contribs 02:48, 11 March 2010 (UTC)Reply

  • My problem has to do with "second hand accounts" and "undergraduate-level textbooks are regarded as tertiary sources". The first is nonsense. The second is not always true, and was worded better already. The text you quoted above is not the problem. --SmokeyJoe (talk) 02:51, 11 March 2010 (UTC)Reply
Okay, that was my copy edit, so if you don't object to it, I'll restore it. SlimVirgin TALK contribs 02:52, 11 March 2010 (UTC)Reply
OK. I'll go away for a while. Simultaneous editing is too hard. --SmokeyJoe (talk) 02:56, 11 March 2010 (UTC)Reply

This is one of Wikipedia's three core content policies. It needs to be stable, and it needs to make sense—both in itself and in terms of its relationship with other policies. When an experienced editor who's adding content is challenged, and he comes here to support his position, he has to know that the policy says roughly what it said when he last checked it. If editors can't rely on it, it may as well not be here.

The problem I've been seeing recently that it has been edited by a few accounts with almost no experience of editing articles, and some of those edits have been allowed to stand for a while. What this does is introduce internal errors and inconsistencies, as well as inconsistencies between this page and the other policies and guidelines. I'd therefore urge editors not to change the policy unless you're 100 percent sure of what you're doing. If you're not, please start a discussion on talk before making or restoring any changes. Cheers, SlimVirgin TALK contribs 03:04, 11 March 2010 (UTC)Reply

The problem is that an inexperienced editor will tweak a good sentence, and will inadvertently change the meaning. This goes unnoticed. Another inexperienced editor comes along, sees the bad sentence, and tries to fix it, but makes it worse. A third editor comes along, possibly an experienced one, and notices only the bad writing; he fixes the writing but doesn't notice the policy is now saying something misleading—and because the writing is now good, the error is less noticeable. And so on until we have a mess. It's no one's fault.
What's needed is for every edit (even minor copy edits) to be made with a full knowledge of what the policy is trying to say, and how it fits with what other policies are saying too. SlimVirgin TALK contribs 03:40, 11 March 2010 (UTC)Reply

Maybe the long discussion above can have some positive effect. Can we change

This includes unpublished facts, arguments, speculation, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position.

to

This includes previously unpublished facts, arguments, speculation, and ideas; and any previously unpublished analysis or synthesis of published material that serves to advance a new position.

This would bring the literal meaning more in line with the intended one. — Carl (CBM · talk) 17:00, 11 March 2010 (UTC)Reply

I don't see what problem adding 'previously' is intended to solve. Previous to what? That only creates ambiguity in a very clear sentence. Crum375 (talk) 18:42, 11 March 2010 (UTC)Reply
The issue is that, since material in a Wikipedia article is published in that article, it's actually impossible for articles to contain unpublished arguments. It's parallel to making a law forbidding people to be inside unoccupied buildings without permission. This is one of two embarrassing gaffes in the existing sentence. — Carl (CBM · talk) 19:11, 11 March 2010 (UTC)Reply
To simply add the word "previously" would introduce an ambiguity. We'd need to rewrite the sentence to introduce the change. People do understand it the way it's currently written, Carl, i.e. "previously" is understood. SlimVirgin TALK contribs 19:18, 11 March 2010 (UTC)Reply
Just as most people would understand the rule forbidding going into an "unoccupied building without permission". Sometimes, more is less. Crum375 (talk) 19:22, 11 March 2010 (UTC)Reply
I agree with SV and Crum here... the "previously" is assumed. Blueboar (talk) 19:28, 11 March 2010 (UTC)Reply
Re Crum: A law forbidding being inside an unoccupied building would be laughable, though. That's the situation we have here. We don't write policies so that people can say, "I already know what that means". We write them so that people who don't already know can learn about them.
The second sentence is just particularly embarrassing because it defines all our writing to be original research (the "new" issue) unless we treat Wikipedia articles as unpublished (the "previously" issue). This is not a question, and I am not asking for an explanation: this is a simple fact about the actual words in the second sentence.
Now it doesn't help to respond, "that's not what the sentence means", because this presupposes that readers already know what the sentence means. I am not asking what the sentence means, and I know what it is trying to say. But new readers don't. For that reason, we need to be careful that if editors simply read the policy language with its plain meaning, they get the right results. That isn't true at the moment, and the second sentence is just a particularly embarrassing example of this. — Carl (CBM · talk) 19:29, 11 March 2010 (UTC)Reply
Re SlimVirgin: I don't see that adding "previously" adds ambiguity; it removes the ambiguity about whether Wikipedia articles count as "published". What ambiguity is added, compared to the lack of ambiguity that makes the current sentence false on its face? — Carl (CBM · talk) 19:29, 11 March 2010 (UTC)Reply
We don't explain what we mean i.e. previous to what? It just introduces another issue that has to be understood without being spelled out i.e. before publication in Wikipedia. We may as well stick with the hidden assumption we have at the moment. Or else rewrite the sentence entirely, but that will bring other issues with it. SlimVirgin TALK contribs 19:33, 11 March 2010 (UTC)Reply
Go for it. — Carl (CBM · talk) 19:34, 11 March 2010 (UTC)Reply
I appreciate the intent of this change, but I'm a little concerned that it will be seriously misunderstood as a requirement that every fact must have been published (at least) twice (once in the source you're citing, and once in some previous source) before it can be included. WhatamIdoing (talk) 21:52, 11 March 2010 (UTC)Reply

Suggestion

Carl, I found it difficult to tweak that sentence alone, so I've rewritten the lead a little. This might also accommodate WhatamIdoing's concern about the need to be clear to new editors. Let me know what you think.

Current lead Proposed lead
Wikipedia does not publish original research or original thought. This includes unpublished facts, arguments, speculation, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position. All material added to articles on Wikipedia must be attributable to a reliable published source, even if not actually attributed in the text. This means that Wikipedia is not the place to publish your own opinions, experiences, arguments, or conclusions.

Citing sources and avoiding original research are inextricably linked. To demonstrate that you are not presenting original research, you must be able to cite reliable sources that are directly related to the topic of the article, and that directly support the information as it is presented. The sourcing policy, Verifiability, says that citations must be added for any material challenged or likely to be challenged, and for all quotations.

"No original research" is one of three core content policies, along with Neutral point of view and Verifiability. Jointly, these policies determine the type and quality of material that is acceptable in articles. They should not be interpreted in isolation from one another, and editors should therefore familiarize themselves with all three.

Wikipedia does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources. It also refers to any analysis or synthesis by Wikipedians of published material that serves to advance a position not advanced by any of the sources who wrote about it.

What this means is that all material added to Wikipedia articles must be attributable to a reliable published source, even if not actually attributed. The sourcing policy, Verifiability, says a source must be provided for all quotations, and for anything challenged or likely to be challenged—but a source must exist even for material that is never challenged. "Paris is the capital of France" needs no source because no one is likely to object to it, but we know that sources for that sentence exist. If no source exists for something you want to add to Wikipedia, it is what we call original research. To demonstrate that you are not adding original research, you must be able to cite reliable published sources that are directly related to the topic of the article, and that directly support the material as presented.

"No original research" is one of three core content policies, along with Neutral point of view and Verifiability, that jointly determine the type and quality of material that is acceptable in articles. Because these policies work in harmony, they should not be interpreted in isolation from one another, and editors should try to familiarize themselves with all three.

SlimVirgin TALK contribs 02:05, 13 March 2010 (UTC)Reply

At the start of the year the Primary sources definition included the following:

"Primary sources that have been reliably published (for example, by a university press or mainstream newspaper)"

This principle has been here for several years. This edit on 11 Jan by SlimVirgin altered it to: "Primary sources that have been reliably published" I can see that and I would support "Primary sources that have been published in a reliable source" but I do not support the total removal of this clause which happened with two more edits the first was this one by TimVickers on 22 January. I understand why the word reliable was removed as a next step because it was meaningless. However the removal of the concept that only primary sources published by a reliable publisher can be used is very important, because it is fundamental to stopping WP:OR particularly in historical articles. If people are free to rummage in unpublished historical archives, they may well be able to overthrow accepted history by digging up a document and quoting an extract from it without violating any other part of PSTS which this sentence was supposed to stop.

So I would like to put back into the policy "Primary sources that have been published in reliable sources," which takes care of TimVickers's comment on his removal "remove 'reliably published', horrible phrase."-- PBS (talk) 13:18, 12 March 2010 (UTC)Reply

BTW I came here to quote PSTS on this because I am looking at several articles were someone has either created a copyright violation or has written a piece of research using unpublished primary sources, (don't know which yet) but removing this clause does not help in stopping Wikipedia being used for OR. -- PBS (talk) 13:18, 12 March 2010 (UTC)Reply

I agree... I think we all agree that Primary sources need to be published (the old manuscript you found in a trunk in your attic is not a good source... as it can not be verified). I hope we would also agree that it needs to be reliably published (the PDF of an old manuscript you found on Joe Blow's website isn't a good source either... Joe Blow is not a reliable publisher). Blueboar (talk) 16:03, 12 March 2010 (UTC)Reply
I can' t see what's wrong with the current text: "primary sources that have been reliably published." To add "primary sources that have been published in reliable sources," repeats the word "sources" (and so is odd writing), and ignores that we use the word "sources" in several different ways on WP. Sometimes it can mean a person, which would make "in" wrong. That's a small point, but I can't see what's wrong with the current sentence. SlimVirgin TALK contribs 22:50, 12 March 2010 (UTC)Reply
Oh, now I see; it was removed. Okay, then I agree that "primary sources that have been reliably published" should be restored. SlimVirgin TALK contribs 22:53, 12 March 2010 (UTC)Reply
In which case it has been reliably published. We consider an expert's website to be reliable for claims as to what the expert says. See WP:SPS Blueboar (talk) 23:35, 12 March 2010 (UTC)Reply
Yes, "reliably published" would include that. SlimVirgin TALK contribs 23:59, 12 March 2010 (UTC)Reply
I have taken the liberty of returning the wording. I have also added the following to the end of the paragraph... I think it helps make the point that material needs to be published (and reliably so) a bit clearer: Do not add material taken from unpublished sources, as that would make Wikipedia the first place of publication for that material. Blueboar (talk)
I think "taken from unpublished sources" is ambiguous, and raises more questions than it answers. I prefer simply "unpublished" or even better, material which has not been reliably published. Crum375 (talk) 23:40, 12 March 2010 (UTC)Reply
Hmmm... The reason why I prefer my language is that I think it ties the caveat directly the concept of NOR. Original research violations stem from what we write about the source, not the source itself. In this case, the specific OR violation is that an editor is writing something based on an unpublished source, not that the source itself is unpublished. So all our statements need to be focused on what the editor is writing (or has written), and less on the source itself. Does that make sense? Blueboar (talk) 23:59, 12 March 2010 (UTC)Reply
The word "source" is used in several different ways on WP: it can be a person, a book, a publisher, or more. This is why I prefer to avoid ambiguity and say what we mean: don't add unpublished material, because that constitutes original research. Crum375 (talk) 00:03, 13 March 2010 (UTC)Reply