Wikipedia talk:No original research/Archive 43

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 40

Archive 41

Archive 42

Archive 43

Archive 44

Archive 45

→

Archive 50

Section "Primary, secondary and tertiary sources" says that Wikipedia itself is a tertiary source. Will this be only the article namespace, or comprise the whole of Wikipedia? For example, can I make changes to articles citing reference desk posts? Jay (talk) 13:18, 11 February 2009 (UTC)

Actually, you shouldn't cite Wikipedia at all ... Wikis are not considered reliable sources. Blueboar (talk) 13:56, 11 February 2009 (UTC)

I think I have got the answer to my first question on article namespace vs talk namespace from Wikipedia:Verifiability#Self-published sources where it says that "Articles and posts on Wikipedia, or other websites that mirror Wikipedia content, may not be used as sources." However, this is only for self-published content on Wikipedia. Can I use Wikipedia posts by others as a source? Jay (talk) 14:19, 11 February 2009 (UTC)

No, because everything in Wikipedia is self-published. Blueboar (talk) 14:34, 11 February 2009 (UTC)

Thanks, I read up the self-publishing article to understand what it is. It talks about the concept as an individual publishing his/her work, and talks about online media like blogs which again are very personalized. The idea of a collaborative medium like a wiki being self-published still seems abstract to me. Can you provide me a reference or discussion which talks about Wikipedia being self published. Jay (talk) 15:40, 12 February 2009 (UTC)

See: WP:SPS. Blueboar (talk) 15:43, 12 February 2009 (UTC)

Thanks, this is the link I mentioned above too. I've finally got it! Examples could have helped. I'll see if I can add some examples for the sub-sections "Self-published sources" and "Using self-published and questionable sources as sources on themselves". Jay (talk) 11:31, 13 February 2009 (UTC)

Translations contributed by by Wikipedia editors would appear to qualify as unpublished facts, arguments, speculation, and/or ideas; and/or as unpublished analysis or synthesis of published material that serves to advance a position. WP:V#Non-English sources appears to make an exception here, however, if no translations published in reliable sources are available. This comes up from time to time in article discussions (see e.g., Talk:Lupang Hinirang#English translation, Talk:Spanish language in the Philippines#Verifiable, neutral and no original research). I suggest that this WP:POLICY article make an explicit exception along the lines of the following:

Exceptions (level 2 header)

Translations (level 3 subheader)

Where English translations of non-English material are unavailable, Wikipedia editors proficient in the foreign language involved may supply a translation citing Wikipedia editors as the source of the translation. Such translations may be challenged on the discussion page of articles involved. If such translations are challenged, editors should cooperate in producing a consensus-supported translation. Translations published by reliable sources are preferred over translations made by Wikipedia editors.

-- Boracay Bill (talk) 02:40, 13 February 2009 (UTC)

Looks good. --Phenylalanine (talk) 02:48, 13 February 2009 (UTC)

The last sentence (about which is preferred) doesn't really address copyright. If the reliable source is not free, then the reliable source is to be preferred for confirming the facts stated in the encyclopedia, but if an actual quotation is used, a free source, such as one contributed by a Wikipedia editor, should be preferred.

This is obvious when it occurs, so I think the simplest option will be to omit the last sentence.

--Hroðulf (or Hrothulf) (Talk) 09:25, 14 February 2009 (UTC)

Typically, a translation of a copyrighted work is still under copyright control of the original copyright holder, as well as control by the translator. So internal translations of non-free text would still be non-free. — Carl (CBM · talk) 15:02, 14 February 2009 (UTC)

Yes, but while there's typically a fair use argument for quoting the original text under discussion, such an argument is less strong when it comes to a translation (because an independent translation would presumably provide as much value to the article). We should probably refer this to the foundation's legal team, but I suspect from a copyright perspective that Hroðulf is right and that our own translation of a source is more likely to qualify as fair use than somebody else's. JulesH (talk) 09:02, 15 February 2009 (UTC)

Of course, it may well be the case that while our own translation would have to be quoted on the page, a reliably sourced translation could merely be referred to in the references section. In this case, the reliable source should clearly be preferred. I think that's covered, however, by WP:NONENG, so need not be discussed here. JulesH (talk) 09:05, 15 February 2009 (UTC)

Carl and JulesH, Unless I badly misunderstand, content contributed by Wikipedia editors is contributed under the GFDL copyright license, and that would include editor-contributed content which is a translation of a foreign text into English. Just to clarify, I'm speaking here mainly of translations in WP articles of non-English quoted snippets which would be fair-use materiel if presented in the original language, but for which a translation is supplied by a wikipedia-editor because no cite-attributable translation could be found in WP:reliable sources. Such a translation would be a derivative work, I suppose, but [fair use]] should apply as much to the translation as it does to the translated non-English text.-- Boracay Bill (talk) 22:02, 15 February 2009 (UTC)

Could some one tell me, if this is the page to seek advice on a problematical distinction between primary and secondary sources, or direct me, if it is not, to the appropriate discussion page? Thank you. Nishidani (talk) 11:00, 15 February 2009 (UTC)

My question will be: if what is described as a 'secondary source', contemporary with the event it describes, happens to be replied to in what is described as a 'primary source', however, written 30 years afterwards by the key actor in the event, how does this affect our definition of primary and secondary sources? The 'secondary source' was written a few months from the event. The 'primary source' written 30 years after the event, by the key actor in that event. The rule given on the relevant wiki page is not clear, and I wish to be directed to the page where informed expertise on this issue is available. I will give full details when I am certain I am on the correct page.Nishidani (talk) 15:41, 15 February 2009 (UTC)

Actually, different accademic disciplines define these terms differently... so we would need to know what subject you are writing about to clearly define what is considered Primary vs. Secondary. But you can use WP:PSTS as a rough guide. Blueboar (talk) 15:53, 15 February 2009 (UTC)

Thanks Blueboar. Here's the gist.

Primary, secondary and tertiary sources The rules are

'Primary sources are sources very close to an event. For example, an account of a traffic accident written by a witness is a primary source of information about the accident. Other examples include archeological artifacts; photographs; historical documents such as diaries, census results, video or transcripts of surveillance, public hearings, trials, or interviews; tabulated results of surveys or questionnaires; original philosophical works; religious scripture; published notes of laboratory and field experiments or observations written by the person(s) who conducted or observed the experiments; and artistic and fictional works such as poems, scripts, screenplays, novels, motion pictures, videos, and television programs. The key point about a primary source is that it offers an insider's view to an event, a period of history, a work of art, a political decision, and so on'

'Secondary sources are at least one step removed from an event. They rely for their facts and opinions on primary sources, often to make analytic, synthetic, interpretive, explanatory, or evaluative claims'

The page where the problem is raised is that dealing with Israel Shahak. The incident is approximately as follows.

In December 1965, Israel Shahak created a stir in Israel by writing a letter to Haaretz complaining that an orthodox Jew had refused him the use of his telephone in order to call an ambulance for an African student he, Shahak, had found unconscious on the street. Use of the telephone would, he was told, constitute a violation of the Sabbath, and in any case Jews were under no obligation to violate the Sabbath except when a Jewish life was at risk.

(a)There was an intensive newspaper coverage of this at the time, which has been analysed by an historian (Segev)

(b) There is an article by a senior rabbi, Immanuel Jakobovits. ‘A Modern Blood Libel--L'Affaire Shahak’ in Tradition, Volume 8, Number 2, Summer 1966, which, in the immediacy of the scandal, was published, a few months after, in an attempt to rebut the charge. This is not an objective review of the incident, but a partisan account by one of the interested parties, at the time, which uses very strong language. I.e.

‘There is nothing more hideous and perverse in the blood-stained history of anti-Semitism than the repeated “blood libels” against the Jewish people . .To substantiate this diabolical libel, the Jew-baiters –usually themselves apostates from Judaism – scanned the pages of the Talmud for out-of-cointext quotations allegedly “proving” the callous disregard of the Jewish religion for the lives of non-Jews... . Havging exhausted tgheir conventional weapons against Jewish traditions, the faanatical jewish secularists have now torn a few pages from the most virulent text-books of anti-Semitism . .The wretched story began with the publication last December of a letter in Israel’s prestigious ‘Haaretz’ by a Dr.Israel Shahak charging that according to “Orthodox” Jewish law it was forbidden to violate the Sabbath to save the life of a non-Jew. To “prove” his inflammatory charge, he subsequently “revealed” that he had himself witnessed an incident in which an Orthodox Jew had refused to allow his telephone to be used to call for help for a non-Jew who had collapsed nearby. Shocked, he asked the rabbinate for a ruling, and they had, so he claimed, confirmed that the Sabbath could indeed be violated only to save a Jewish life.

These “revelations” aroused a storm of public indignation in Israel and throughout the Jewish world. Editorials, suitably embellishing the story, fumed against the “medievalism” and “the utter absurdity and inhamnity” of the rabbis. In all thie excitement it was almost overlooked that Dr. Shahak, challenged to substantiate his personal “testimony”, was eventually forced to admit that the Orthodox Jew he had “witnessed” refusing the use of his telephone simply did not exist. The whole incident had been fabricated . .Equally overlooked was the circumstance that the rabbinate, far from having confirmed Dr.Shahak’s allegation, had in fact ruled that the Sabbath must be violated to save non-Jewish no less than Jewish lives’ (Jakobovits's article p.59)

(c) There is Shahak’s own version, three decades latter, Jewish History, Jewish Religion: The Weight of Three Thousand Years (1994). ‘I had personally witnessed an ultra-religious Jew refuse top allow his phone to be used on the Sabbath in order to call an ambulance for a non-Jew who happened to have collapsed in his Jerusalem neighbourhood. Instead of simply publishing the incident in the press, I asked for a meeting with the members of the Rabbinical Court of Jerusalem, which is composed of rabbis nominated by the State of Israel. I asked them whether such behaviour was consistent with the interpretation of the Jewish religion. They answered that the Jew in question had behaved correctly, indeed piously, and backed their statement by referring me to a passage in an auythoritative compendium of Talmudic laws, written in this century. I reported the incident to the main Hebrew daily, Haaretz, whose publication of the story caused a media scandal. The results of the scandal were, for me, rather negative. Neither the Israeli, nor the diaspora, rabbinical authorities ever reversed their ruling that a Jew should not violate the Sabbath in order to save the life of a Gentile. They had much sanctimonious twaddle to the effect that if the consequence of such an act puts Jews in danger, the violation of the Sabbath is permitted, for their sake.’ p.1

(d) We have one indubitable secondary source recounting the incident by the Israeli historian Tom Segev in his, 1967: Israel, the War, and the Year that Transformed the Middle East, tr. Jessica Cohen, Macmillan, London, 2007, pp. 99-100.

I take Segev to be an ideal secondary source. His version differs from both Shahak and Jakobovits, and directly narrates the newspapers of the day that covered the news. Shahak’s account, though 3 decades later, does fit the simple definition of a primary source. I have a problem with classifying Jakobovits’s article. It is ‘one step’ from the event, (secondary source) in that it was written a few months (early-mid 1966) after the scandal broke (December 1965). However, it was also written ‘very close to the event’, by an interested party to the matter (a chief Rabbi in England writing on behalf of his Israeli colleagues), makes many remarks that cannot be verified by other sources, and fits the description: ‘The key point about a primary source is that ‘it offers an insider's view to an event, a period of history’, a work of art, a political decision, and so on.’

Jakobovits’s version is evidently an insider’s account since it asserts many things that can be neither found in Shahak’s account, nor in the historian Tom Segev. Jakobowitz is a contemporary witness for things he asserts on his own authority to have occurred, but for which there is no independent corroboration in available historical sources.

The second problem is that Shahak, who is taken to be a primary source, wrote his account long after this so-called secondary source, and explicitly alludes to Jakobovits’s account and the evidence Jakobovits used. Can we define Shahak’s book as a primary source, when its account refers to what some define as a secondary source that was written three decades before his own account? If so, we would have the curious situation in which a source by a third party written close to an event (1966), is regarded as ‘secondary’, while a source written bv a first party (Shahak 1994), written long after the event, and in response to the earlier source, is regarded a ‘primary source’, though it is farther from the event than the other source.

I think Jakobovits is a borderline case, and doesn’t fit neatly the definition of a secondary source. I don't know how to qualify Shahak's own book, which is 'primary' but written after, and in reply to, a source others regard as 'secondary'. Confusing.

Thanks for your attention. Nishidani (talk) 17:20, 15 February 2009 (UTC)

Jakobovits, an authority on Jewish law, wrote an account of Shahak's letter and the allegations contained in it, and published this in a 3rd party, reliable source, months after the event. He analyzed the alleged incident, the claims by Shahak, and counter claims. He is neither an "insider" (not having been present to witness the incident, nor even in the same part of the world where it occurred), nor is he a "partisan" - he is a published expert on the matter at hand. He is clearly a secondary source, just as Segev is, and there is nothing remotely borderline about this. NoCal100 (talk) 18:08, 15 February 2009 (UTC)

I hope to get input from parties outside of this dispute, neutral to the matter, and analysing the issue in purely formal terms. So I will withhold any comment I might be tempted to make until we have a variety of opinions. I will only note that the passage in question, or which I questioned, is a synthesis of at least three sources, excluding Segev, but is/was sourced only to Segev. That itself is the gravamen of my complaint. Wrong attributionNishidani (talk) 18:46, 15 February 2009 (UTC).

If you want to get neutral input, you need to avoid biased framing of the case, such as "Jakobovits’s version is evidently an insider’s account", or "This is not an objective review of the incident, but a partisan account by one of the interested parties" NoCal100 (talk) 20:11, 15 February 2009 (UTC)

Not everybody makes the distinction but in my opinion which I think some share, the status of a source as primary or secondary depends on what it is being used for. Citing Shahak's letter as a source for the events described in he letter is secondary sourcing, albeit most likely unreliable (a first-person account, written in circumstances that do not indicate fact checking or editorial oversight, etc). Citing the letter as a source for the contents of the letter is primary sourcing. You are looking at a document and informing the reader of the contents of the document, which comes fraught with all of the issues that come with original analysis of texts. Citing Jakobovitz for the content of Shahak's letter is a secondary source - whether or not it is a reliable source is beyond the scope of the original question. Maybe the reliable source noticeboard WP:RS/N is a good place to ask these questions but it's sometimes hard to get a good discussion going there, particularly if people start disputing things. Wikidemon (talk) 20:51, 15 February 2009 (UTC)

In general, if two sources dispute facts, the way to handle it to mention both... this in essence shifts the the article from discussing satements of fact, to discussing statements of opinion... mentioning what both sources say on the issue, and attributing the various viewpoints to their authors. Blueboar (talk) 21:01, 15 February 2009 (UTC)

Wikidemon. Shahak's letters to Haaretz are not cited, and are not known. We only have his retrospective account of the affair on page 1 of his 1994 book, 29 years after the event, and 28 years after Jakobovits' version. The problem is that the passage was written from three sources (Shahak, Jakobovits and someone else) and then sourced to one other source (Segev). Trying to tell the editors they have screwed up, and to rtewrite this simple couple of lines according to all three sources, has created an edit-war, and the introduction of one extremely poor source, with an WP:OR infraction added, to boot. Since I am apparently disliked, antipathy is governing this, not rational judgement. The solution is simple. Readjust the text to the 3 RR. No deal. I raised the primary/secondary issue because Shahak's own version is contested by Jakaobovits, but Shahak wrote in hindsight of Jakobovits's account, after 3 decades, taking it into account. Since it can fit the def. of a 'primary source', the end result, if Jakobovits is a 'secondary source' is that in editing the account Jakobovits has authorial priority, even though he was involved in the immediate aftermath of that event, and is as partisan as the primary source.Nishidani (talk) 22:08, 15 February 2009 (UTC)

Without weighing in on the specifics (there seems to be a "pro" and "anti" Israel thing going on here, and I haven't bothered to figure out which side is which), it looks like there may not be any reliable sources at all. A primary source is problematic in one way, a partisan secondary source in another, and there's no firm rule that one is better than the other. If the only references you can find to a particular letter are the author's recollection of it decades later, and a couple partisans arguing over it, maybe the matter just isn't notable. The solution to conflicting weak source isn't always to mention them all. Sometimes the best conclusion as a matter of WP:WEIGHT is that if no neutral reliable sources have considered the matter significant enough to write about, it is not Wikipedia's role to bring it up either.Wikidemon (talk) 22:21, 15 February 2009 (UTC)

Nothing anti-Israel in my editing, as opposed to how it is perceived. I am trying to get a page by a Jew, and an Israeli of great distinction, written according to reliable sources. Tom Segev is an historian, and he recounts the episode in his book. He is an undoubted reliable source. He was cited for information that however was written up tacitly from the other two sources, Shahak and Jakobovits. It was this misuse of Segev, an RS by an historian, to document a different account from the one he gives. The episode has to be mentioned, since it was significant.Nishidani (talk) 22:31, 15 February 2009 (UTC)

I'm not accusing anyone of bias. But how do you know the episode (the "misuse") is significant? Wikidemon (talk) 23:27, 15 February 2009 (UTC)

Segev was 'misused' in that his account differs from Jakobovits' and Shahak's, yet, while they were used to write the passage, Segev was used to source the passage. Nishidani (talk) 12:45, 16 February 2009 (UTC)

So, there's a bizarre little flap at Talk:Toa Payoh ritual murders about whether it's permissible to link or otherwise identify a "Siamese sex god" described in Singaporean police reports as "Phragann" or "Pragnang" to the apparent correct spelling, Phra Ngang (Thai พระงั่ง). (Both spellings are transcripts from interviews conducted between non-Thai people; neither spelling is phonologically possible in Thai; they do not appear anywhere on the net outside this article; and no alternative identifications of either have been proposed.)

User:Jappalang is of the opinion that, since the police report and various books relying on said reports do not use this spelling, it would be "synthesizing an original theory" to even hint at this in a footnote. To me, this sounds like claiming that, if a source uses the name "Misissippi River" for a large river in Midwest and Southern US that forms a border line for 10 states, it would be "original research" to identify it as the Mississippi River. Third opinions? Jpatokal (talk) 09:09, 16 February 2009 (UTC)

As far as I know this policy applies only to verifying factual statements made in article space. It does not apply to style issues such as how to spell things.Wikidemon (talk) 10:27, 16 February 2009 (UTC)

After looking at the discussion I am afraid it's not as easy as that. Think of a US source claiming that "the guv'nor killed him!" By regularising the orthography you will reinforce the reading that it is the governor of the state in question who is being accused (rather than someone's boss or landlord). Unless you have other sources which imply that this is in fact the intended meaning, it's probably improper synthesis to regularise. I am not sure about the present case, but it seems to be similar. --Hans Adler (talk) 12:20, 16 February 2009 (UTC)

I do not follow that. That would be a simple mistake of misreading the source, not conducting original research on it. We run into that all the time, more or less anytime we substitute a different word for one in the source. Wikidemon (talk) 12:41, 16 February 2009 (UTC)

A glance at Wiktionary (or your favorite dictionary) shows that "guv'nor" is an actual word known to have a separate meaning from "governor", and the two are clearly not equivalent This does not appear to be the case for either of *phragann or *pragnagn.

Also, I'm fine with keeping the original spelling (in quotes) in the article body, but I think it does the reader a disservice not to identify the correct spelling in at least a footnote. Jpatokal (talk) 15:51, 16 February 2009 (UTC)

Bump. Any more opinions about this? Jpatokal (talk) 03:03, 19 February 2009 (UTC)

There is a whole other issue in "correct spelling" when foreign names and terms are involved. Is it correct to write "Franz Josef Strauss" or "Franz Josef Strauß"? The naming of pages is addressed in WP:UE and the naming within a page is addressed at WP:MOS#Foreign terms. Also see Wikipedia:Avoid neologisms --PBS (talk) 22:21, 21 February 2009 (UTC)

Yes, that's another kettle of fish, but that has nothing to do with this particular case, where the correct Thai romanization is clear. Jpatokal (talk) 03:30, 22 February 2009 (UTC)

So after my previous engagements on this page, I decided to sit and look about some more to get a better sense of the issues I'm having.

Today, I ran into something that very much renewed my sense that this page, in its current form, is damaging the encyclopedia. Resident Evil 5. [1] is the version in question, in case this has been fixed. Scroll down to footnotes 43 and 44, in the "accusations of racism" section.

The information those are being used for - N'Gai Croal's critique of the game - come from a web-published source. It's a reliable source - Croal is a significant figure in the field, and the interview it comes from was with MTV.

But because we prefer secondary sources, instead of linking to Croal's comments (which are easily found) we link to secondary sources that quote them selectively. That is to say, we opt to link to sources that give a less complete view of Croal's position, even when a more complete source is readily available. Why? Because we prefer secondary sources.

Why do we do this? Presumably because we wanted the secondary sources to show that Croal's comments were significant. Because people objected on the talk page. (And if you look on the talk page, you'll find plenty of fans of the series who dislike the racism section because they disagree with Croal) But in doing so, we let sourcing stop being a service to the reader and start being a service to the talk page.

Any reader who is interested in following a source for a claim about what Croal said will be better served by a link to Croal. Other links may serve any number of purposes. But from a reader perspective, the primary source is simply the more helpful source there.

We need to distinguish between what we use to make editorial decisions and structure the article and what we present as information and as sources. As it stands, this policy encourages writing articles in a manner that is less helpful to the reader. Phil Sandifer (talk) 21:42, 18 February 2009 (UTC)

How did this policy enter in the decisions made in the article? Professor marginalia (talk) 00:32, 19 February 2009 (UTC)

the preferences for secondary sources (as I understand it) is based on the idea that secondary sources are not engaged in promoting a position or belief the way that primary sources are. for a silly example, say Sir Isaac Newton comes along and says 'here's is a theory that explains all motion in the universe!'. he is the primary source for that theory, and emotionally invested in having other people accept the theory as true. later (secondary) sources who talk about or use his theory are (assumedly) not personally attached to the theory, but only concerned with whether it's useful. secondary sources, thus, give both emotional detachment and historical perspective. the problem, though, is that the primary/secondary is hard for a lot of people, and sometimes hard even for people who are well-practiced with it. in this case, for instance, the Godinez article is (I think) bordering on primary research (because of Godinez' admissions of his own sympathies), but would probably pass as secondary given that he does put some effort into discussion the issue from different perspectives. I also think that he's been badly quoted in the article (that quote's been cherry-picked for minimal impact - you could probably expand it some to make it closer to what Croal actually said). further, news interviews are rarely considered to be primary sources, since the news person (if s/he's doing the job correctly) is not editorializing, so you could probably justify using the Croal interview directly. but that's a thing to be argued out on the talk page. --Ludwigs2 01:03, 19 February 2009 (UTC)

I think that idea is an overgeneralization, though. Depending on the source, secondary sources may be just as, or more, vested in a particlar ideology or belief than the primary source. The classic example is the field of creationism. Which is more reliable and neutral: (1) the primary scientific findings of a microbiologist or paleontologist in a scientific journal, or (2) secondary commentary about those findings by creationists and evolutionists?

Primariness/secondariness is a very crude "gauge" to measure reliability or neutrality of a source, and I agree with Phil Sandifer that sometimes this policy "guestimate" results in real harm to articles. COGDEN 01:34, 19 February 2009 (UTC)

Well, but I think a larger issue is that we should distinguish between what we use for behind the scenes editorial decisions and what we use as sources to present information. In this case, the secondary sources establish the importance of Croal's comments, and justify paying significant attention to them. They are used to make the editorial decision to include Croal. In that sense, our coverage is "based" on secondary sources. However our actual coverage - what we present to readers - is better done if we use Croal. And that's an important distinction - what is best for decision-making is not always best for presentation. Phil Sandifer (talk) 01:47, 19 February 2009 (UTC)

All I could find debated on the talk page was whether or not Croal's remarks in the MTV blog were notable. Once other sources were found demonstrating that there was quite a fuss made about Croal's remarks the notability question seemed settled. The article reads, "Resident Evil 5's 2007 E3 trailer was questioned for its depiction of a white protagonist killing black zombies in an African village. Newsweek editor N'Gai Croal began the criticism, stating, 'There was a lot of imagery in that trailer that dovetailed with classic racist imagery.' He acknowledged that only the preview had been released." The original interview can't be used to source the claim, "Croal began the criticism." Strictly speaking, I don't believe the secondary source made this claim exactly. I believe that source wrote though that Croal's interview led to a bigger whoopla than did the earlier stir up at Electronic Entertainment Expo. I don't see that anyone objected to citing to the interview itself. Having read the original, I don't find the quote or Croal misused by Gobinez. There might be a point made that the wp article didn't represent what Gobinez or Snow wrote as well as it should do. I think this is a tempest in a teapot frankly, one that has nothing to do with the original research policy. I found no reference to it on the talk page. The secondary sources are good, actually, attempting to describe the heated controversy that followed in the wake of Croal's remarks. The primary source can't cite any such claims about the reactions to itself. The description there in the wp article could be improved to better represent what these secondary sources wrote-it wouldn't be hard. Cite the interview as well. It's not a big deal. Professor marginalia (talk) 01:50, 19 February 2009 (UTC)

@phil: just as a clarification, citing other sources does not make a source secondary. almost all primary research cites previous research (in order to refute it, or to ground its own claims). secondary sources are distinguished by not advancing a particular viewpoint of their own, in the creationism case, creationists are as likely as microbiologists to be involved in primary research (advancing a truth claim over other truth claims). part of the problem with creationism, in fact (and the reason it's generally considered pseudo-scientific), is that creationists are willing to advance claims of primary research, but unwilling to allow those claims to be examined and critiqued in the ways that are normally expected of primary research. the whole debate kind of like that old Monty Python skit with Anne Elk, actually...

I think we are discussing two seperate issues here. The first issue is what is the best source for verifying that someone said what we, in the article, say they said. I think that for a statement as to someone's opinion the best source will almost always be the source where the opinion was stated. We should cite the original, rather than relying on a second hand quote passed through a secondary source. This is one of the main reasons why we DO allow primary sources. So, if we are going to directly quote what Croal said in the MTV intervew, the best source is that interview.

The second issue is demonstrating that the author of the quote (and thus his/her comments) is notable. For that, we need to show that some other than the author cares about what he or she said. We need a secondary source that has taken note of what the author said.

In other words, if there is a conflict here, it is between WP:V and WP:NOTE. My suggested solution to that conflict is: cite both the MTV interview and the newspaper report on it.

What I don't understand is how any of this relates to WP:NOR... is anyone arguing that there is some sort of NOR violation going on in the article? Blueboar (talk) 03:28, 19 February 2009 (UTC)

It doesn't relat to original research at all, but historically, this discussion about primariness/secondariness has taken place here, under the theory (dubious, imho) that citations to primary sources are more likely than other citations to involve original research. That heuristic, however, really goes afar afield of the policy behind NOR, and is like using a jackhammer to solve a probelm better solved with a scalpel. We all know that some uses of primary sources are bad, but so are some uses of secondary sources. What makes them bad is not their primariness or secondariness, but how the citaiton is used, and for what purpose. Primary sources are the most reliable sources as to what the primary sources say. Secondary sources are more reliable (but not infallible) at providing context. COGDEN 19:09, 19 February 2009 (UTC)

Yes, there is a place for primary sources: they work very nicely for undisputed routine facts--where someone went to school, the length of a movie, who the characters are, the plain description of what happens in the plot. But if we are to describe a plot, usually the best source is to carefully watch the film or other work, rather than to rely on a critic. A critic typically discusses selected elements of the plot to make the desired point point; a secondary article on the work selects the material to support its hypothesis. Sometimes there are reliable secondary sources that do discuss the work in an encyclopedic fashion, they should always be searched for, and then they can and should be used for this. Of course, for anything involving interpretation, we need others to do the interpreting. This is not a hard and fast distinction, because some interpretation is necessary in picking out what facts to use and exactly where to get them. DGG (talk) 13:06, 21 February 2009 (UTC)

One OR-like issue that comes up with primary sources is determining their relevance and weight. For example, it often happens that a well known person, say a politician, journalist, pundit, etc., will make a statement about something. If a secondary source covers that statement we have some indication that it is (small n) notable, i.e. that reliable secondary sources consider the matter significant enough that their readers should be informed of it. However, if the only citation for what the person said the place where the person said it, e.g. a book, blog, radio interview, etc., then it is original research to use that source alone to verify that the statement was made. In the case of watching a film to describe film plots that is not usually questioned. In the case of reading a book or a blog in order to add controversial claims to the encyclopedia it is a frequent problem. A statement like "Historian X claimed that Margaret Thatcher was the most inept prime minister in modern British history" could be unquestionably true if cited to a book where historian X said such a thing. But it could be misleading (say, he later retracted the claim or came to change his mind), or utterly trivial (X was a historian of little note, or represents a fringe viewpoint, or is on the payroll of a partisan anti-British organization, or does not specialize in modern British history and as such is talking outside of his area of expertise). We trust reliable secondary sources to evaluate whether something is worth repeating, and to investigate and explain the context for us. When we go straight to the source we can get sidetracked, and attach undue weight to things we think are important but are really not. Wikidemon (talk) 21:46, 21 February 2009 (UTC)

I notice that there may be at least one problem with the following example in the WP:SYN section of WP:NOR. I highlighted in bold the main parts pertaining to the problem.

---

Here is an example from a Wikipedia article, with the names changed. The article was about "Jones":

Smith says that Jones committed plagiarism by copying references from another book. Jones denies this, and says it's acceptable scholarly practice to use other people's books to find new references.

That much is fine. Now comes the unpublished synthesis of published material. The following material was added to that same Wikipedia article just after the above two sentences:

If Jones's claim that he consulted the original sources is false, this would be contrary to the practice recommended in the Harvard Writing with Sources manual, which requires citation of the source actually consulted. The Harvard manual does not call violating this rule "plagiarism". Instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them.

---

It's not clear whether Jones claims "that he consulted the original sources". He only refers to using "other people's books to find new references." I would suggest the following rewording at the beginning of the last paragraph of the example.

If Jones did not consult the original sources, this would be ...

--Bob K31416 (talk) 17:19, 12 February 2009 (UTC)

I agree. This would be an improvement. --Phenylalanine (talk) 02:26, 13 February 2009 (UTC)

There are a few problems... the first two sentences are not really fine. The second is a run-on that is itself synthesis or original research by implication. Jones denied Smith's claim and he said something about scholarship. To be valid sourcing, the source would have to verify that Jones' statement about valid scholarly practice was in fact a denial in response to Smith's accusation. Also, the problem with the later is analysis more than synthesis: it is use of a general principle to advance an argument about a specific sourced thing, a narrow type of synthesis. A simpler and more general example of synthesis would be the combining of two sourced things, e.g. "Smith did X.[cite] However, Smith later did Y,[cite] which seems to contradict X[uncited]. As a further objection, narrow examples are usually not a good idea in policy sections. Couldn't this be explained more succinctly and forcefully by saying it directly rather than telling a story about editing process, how editors piled bad content into an article? Wikidemon (talk) 15:33, 14 February 2009 (UTC)

I made some changes re the first two sentences of the example and context, just before you posted your message. Here's what it looks like now:

Here is an example that is based on an excerpt from a Wikipedia article, with the names changed. The article was about "Jones":

Smith says that Jones committed plagiarism by copying references from another book. Jones denies this, and says it's acceptable scholarly practice to use other people's books to find new references.^[27]

That much is fine, since it is directly supported by a reliable source.

--Bob K31416 (talk) 15:59, 14 February 2009 (UTC)

Well, I've blustered through and rewritten the whole example. No offense intended but there were so many issues with weak language, and so much that was besides the point. Starting with the first sentence, the example is not based on an excerpt the way it is written, it is based on an incident. Pointing out the name change is both irrelevant and redundant (done twice, once directly and the other by the self-conscious use of quotes), and in fact it is not pertinent to the example what the article is about. Then in the supposed excerpt, the phrase "another book" suggests that there is a first book...but the subject of the sentence is an author, not a book. It equates an author and a book. And so on. I tried to write everything to be very literal and straightforward. Wikidemon (talk) 16:17, 14 February 2009 (UTC)

My edits just got reverted for no reason I can see other than being too bold.[2] Do I need to enumerate everything that's wrong with the example? It's probably 20 things.Wikidemon (talk) 16:32, 14 February 2009 (UTC)

Actually, there was a reason given and a suggestion too, which brings us here to this talk page.

I think some of your ideas have good potential. Let's look at them one at a time. OK?

Re your remark - "Starting with the first sentence, the example is not based on an excerpt the way it is written, it is based on an incident."

Could you clarify your point? Thanks. Here's the sentence for reference:

Here is an example based on an excerpt from a Wikipedia article, with the names changed. The article was about "Jones":

-Bob K31416 (talk) 16:52, 14 February 2009 (UTC)

Piece by piece:

• "...based on an excerpt..." (the one you ask about). This introductory sentence promises the reader that the prose to follow will be an example based on an "excerpt" from a Wikipedia article. However, the next few sentences are not precisely what the introduction promises. They recount that there was an article with a couple sentences that were "fine", but then someone added some more sentences, which are not fine for reasons explained. That narrative describes an editing process, the addition of a policy violation to a formerly compliant article. Thus it is an event, a passage of text, that the example is based on. The reader can easily forgive the imprecision because the intent is clear, but in so doing it creates a slight cognitive dissonance, and suggests that this policy page is loosely written - something that weakens its authority. It is more correct to say that the example is based on an "event" or "incident", but that calls undue attention to an irrelevant matter. So better to use the more neutral wording that it is based on an "article", because as a general term, an article implicitly includes the history of the article. Ideally we would find a more neutral, literal way to introduce the example because it does not matter what the example is based on. That is impertinent.
• "Here is an example..." is an informal tone that implies a narrator, as opposed to being more neutral and authoritative, e.g. "The following example..."
• "Here is an example..." followed by the example. Pointing out the obvious / telescoping / redundant. It's all surplus text, better get to the point.
• ...with the names changed...about "Jones". That says twice something that does not need to be said at all, that the article uses fictitious names.
• The article was about "Jones" - The statement is self-contradicting in a convoluted way. The article is not about Jones. Jones is a fictitious name. The article is about a person with a different name that has been changed, presumably to avoid indirectly identifying a Wikipedia editor whose purported act of bad editing is enshrined here in policy. But the subject of the article is immaterial - it does not affect the example.
• All in all, it is a weak way of introducing an example. We don't need to say it is an example (that is obvious and redundant), we do not need to name the subject of the article because we can't, we don't need to explicitly describe the process of changing the names to protect the guilty (that is administrative overhead that could be stated far more simply if necessary). The only thing we are really trying to say is something like "The following is based on a Wikipedia article". But even that is dubious because the example has been changed, and there is nothing to be gained by convincing the reader it was a real-live Wikipedia event rather than something made up for this policy page - "true story, only slightly fictionalized" does not give the example any more authority.

That's what I have for the introductory sentence. One way or another the English will be a lot better if it is simplified and shortened by half or so. Wikidemon (talk) 18:13, 14 February 2009 (UTC)

[Please note that this message of mine is slightly out of chronological order.]

I think you made some good points. How's this,

The example in the following discussion is based on some editing of an actual Wikipedia article.

--Bob K31416 (talk) 21:11, 14 February 2009 (UTC)

Just a reminder that the original insertion of WP:SYN on 11 April 2006 can be found here. The text of the then-newly-placed section can be seen here. After basic copyediting, it read like this on 12 April 2006. The original example of text asserted to be OR, as it was first used in the section that is now WP:SYN, is:

If Jones's claim that he always consulted the original sources is false, this would be contrary to the practice recommended in the Chicago Manual of Style as well as Harvard's student writing manual, both of which require citation of the source actually consulted. Neither manual calls violations of this rule on citing original sources "plagiarism." Instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them.

This paragraph, along with the one above it, was intended to be a direct quote from the article's editing history, with Smith and Jones substituted for Dershowitz and Finkelstein. Offhand I think a record of it should be preserved on a conveniently accessed archive link if WP:SYN is to continue to use what was intended to be a direct verbatim example from an actual article edit. The Dershowitz example, and the wording generally, is also discussed in Wikipedia talk:No original research/Archive_38, sections 9-13. ... Kenosis (talk) 20:46, 14 February 2009 (UTC)

Additional note: For convenience, I'm copying and pasting here a couple parts of another post I previously made, now in Archive 38: The original editing disagreement to which the example refers appears to be this, amid a number of edits leading up to as well as following the one I link to. The article text, in the section on Dershowiz' response, stated:

If Dershowitz's [Jones's] claim that he always consulted the original sources is false, this would be contrary to the practice recommended in the Chicago Manual of Style as well as Harvard's student writing manual, but neither of these sources call it "plagiarism."

In the original source from which this edit is derived, Finkelstein ("Smith") says: "It is left to readers to decide whether Dershowitz ("Jones") committed plagiarism as defined by Harvard University -- "passing off a source's information, ideas, or words as your own by omitting to cite them." (Here's the link.) So "Smith" never used the words ""cit[ing] the source consulted", but rather, it appears the additional language was a WP user's contribution, thus resulting in a synthesis consisting of material from both the Harvard and Chicago manuals on this project page. There might be something vaguely ironic about that in the context of the WP:SYN section of WP:NOR, except that this is the WP policy page and not a WP article.
.....Admittedly this can all be fairly confusing. Possibly in the end this example will be a bit less confusing more stable to use another example that doesn't involve a quote of an actual edit, one that isn't quite as complicated as this one? Perhaps one that doesn't involve the additional element of defining plagiarism? I don't happen to have at hand a suitable replacement for this example at the moment.. ... Kenosis (talk) 21:01, 14 February 2009 (UTC)

This is very useful info. Perhaps it should be in its own subsection or section? So far I have read it once without looking at the links, but I will spend the time to carefully study it. Perhaps a consideration of these historical versions of the present example will help make it better. --Bob K31416 (talk) 21:30, 14 February 2009 (UTC)

We could footnote the introduction to link to the article or its edit history. But then, was that a watershed in the understanding of OR, a particularly epic editing dispute, randomly chosen? An example of SlimVirgin changing policy to suit her opinions?[3] I wonder if we should use this example at all. It's distracting to bring up the Dershowitz–Finkelstein affair, a topic of edit warring that continues today and is covered by Israel/Palestine general sanctions, and then a little peculiar that the names were changed to anglicize them. Also, Ragout (talk · contribs · deleted contribs · page moves · block user · block log), the editor who was edit warring in favor of the right version on that article,[4] was a likely WP:SOCK who showed up, immediately began edit warring claiming to be fixing POV violations on two subjects, edit warred against SlimVirgin on this policy page, accused SlimVirgin (talk · contribs · deleted contribs · page moves · block user · block log) of wikistalking, then disappeared from Wikipedia after fewer than 500 edits. That's all too much Wikidrama. The inclusion of that specific example smacks of agenda-driven policy building. Better to make the example generic or find something uncontroversial like OR on an article about tea and scones.Wikidemon (talk) 22:58, 14 February 2009 (UTC)

Wikidemon, Please note that I responded to your message about the intro sentence, with a proposed edit which hopefully addressed some of the issues you mentioned. Could you look at it? Thanks.

(P.S. Your last message looked like a digression into personal issues. I would suggest staying on topic.) --Bob K31416 (talk) 23:57, 14 February 2009 (UTC)

Yes, thank you. I like your version of the introduction. I was commenting on whether or not it could or should include a footnote linking to the actual incident. The original research example included in the section indeed brings up some issues, but I wouldn't call them personal. That's on topic because I was answering my own question, whether there is a good reason to use this particular example. In a quick review of the matter I found residue of an old WP:BATTLE. If it is just an artifact of questionable editing, I don't think we need to give it any special deference -- we are better off rewording it to most clearly illustrate the problem of original research for the reader than preserving it in its original form as a piece of Wikipedia history. Even a footnote would be unnecessary, if that only serves to direct readers to murky old edit wars on controversial topics. Wikidemon (talk) 00:21, 15 February 2009 (UTC)

Thanks for supporting my proposed version of the intro sentence, which, I must acknowledge, came about because of your comments. : )

Would you like to continue the discussion with another one of your points?

(P.S. As far as the historical versions are concerned, as I mentioned I will be studying it. I will try to see if there are any ideas in it that we can use. I can't tell if it is useful yet without studying it first.) --Bob K31416 (talk) 00:36, 15 February 2009 (UTC)

Sure, thanks for looking into it. You seem thorough, deliberate, and fair-minded on the subject so I'll trust you to get to the bottom of things. I'll see if there is anything else I can do, but there is a preliminary question that applies to most of the material, whether to the same example, and if so, how much liberty can we take to simplify it for purposes of clarity.Wikidemon (talk)

Thanks. I changed the intro sentence in our wiki. I added a sentence about using Jones for the name of the person who is the subject of the article since I thought that might be helpful in understanding the example. OK? --Bob K31416 (talk) 00:52, 15 February 2009 (UTC)

After further research, I suggested a different change related to the example's article subject. Please see the following subsection. --Bob K31416 (talk) 16:42, 15 February 2009 (UTC)

Subject of article mentioned in WP:SYN example

The subject of the article mentioned in the version of the example[5] that I just changed, was not one person but a dispute between two people.[6] The example calls the two people Smith and Jones. I made the following change:[7]

from - In our example, the person who is the subject of the article will be called Jones.

to - The subject of the article was a dispute between two people who will be called Smith and Jones.

--Bob K31416 (talk) 16:36, 15 February 2009 (UTC)

From a narrative standpoint I don't like the redundancy or use of future tense. How about this: The following example is based on some editing of an actual Wikipedia article about a dispute between two authors, "Smith" and "Jones" (names changed for purposes of illustration) - Wikidemon (talk) 22:35, 15 February 2009 (UTC)

Hi. Here's what it currently looks like:

The example in the following discussion is based on some editing of an actual Wikipedia article. The article was about a dispute between two people who will be called Smith and Jones.

Re redundancy (the use of the word "article" twice?), I suppose I could change it to:

The example in the following discussion is based on some editing of an actual Wikipedia article about a dispute between two people who will be called Smith and Jones.

but I think the sentence would be too long for good reading.

Re part of your suggestion, "a dispute between two authors, 'Smith' and 'Jones' (names changed for purposes of illustration)" - this seems too complicated with quotes and the phrase in parentheses, in my opinion. I tried to avoid these problems by using "will be called Smith and Jones", which seemed the simplest and clearest way to express the ideas. However, changing "people" to "authors" seems like an improvement.

Also, I don't see that the names were changed "for the purposes of illustration." The example could have been just as well illustrated by the original names, but that would have been a distraction, in my opinion, since the example is meant to be about improper synthesis, rather than Dershowitz and Finkelstein, who the reader may know from TV etc, especially Dershowitz.

Re part of your suggestion, "The following example" - I thought you objected to this before because what followed wasn't only the example but the example intermingled with discussion. That's why I used the phrase, "The example in the following discussion" to address what I seem to recall was a concern of yours in a previous message.

--Bob K31416 (talk) 23:41, 15 February 2009 (UTC)

Well, I guess the tone is still a little pedantic, and includes red herrings that are artifacts of the fact that we've chosen (so far) to use as an example a piece of editing history that raises other issues and is perhaps too controversial to describe exactly as it happened. Just why are we using this example, and why are the names changed? If our only purpose is educating the reader about synthesis we could just say "Consider the following example:" Wikidemon (talk) 23:52, 15 February 2009 (UTC)

Re "If our only purpose is educating the reader about synthesis we could just say 'Consider the following example:'" - That came to my mind too but I felt that the reader would feel that it is more relevant and worthwhile if it was based on an actual piece of wiki editing. BTW, I changed "people" to "authors". --Bob K31416 (talk) 23:58, 15 February 2009 (UTC)

Next narrative segment

Regarding the next section of narration...

That much is fine, since it is directly supported by a reliable source. Now comes the unpublished synthesis of published material. The following material was added to that same Wikipedia article just after the above two sentences:

...I have the following objections / concerns:

1. That much is fine.. - no, actually, it isn't. It's rather weak writing, something I'll address later. We shouldn't express approval, nor does that matter for purposes of the illustration. How about being more specific: "This much is not original research" or "That much does not violate this policy" or the like?
2. since it is - the word "since" refers to relative timing. Until recently it was considered nonstandard usage for "since" to introduce a second statement that provides a logical justification for a statement just made (A is true since B is true). That has gained some legitimacy through usage but is still weaker English than using "because" (A is true because B is true). Note that "because" is also not quite correct, in that "because" originally referred to causation, not deduction. But it is nonetheless more proper than using "since".
3. directly supported - "Directly" is an empty filler word here. It has some intuitive appeal because we're trying to distinguish between matters that are contained in the sources (i.e. "directly") versus matters that are deduced by combining multiple sources (i.e. synthesis). But the opposite of direct is "indirect", not synthesized. Actually, that's three points on a triangle: directly supported statements, indirectly supported statements (i.e. it has to pass through a second logical step), and synthesized statements (two sources are combined to support one thing). It's not a slam dunk but I would delete the word "directly" because it makes the reader question what it means for something to be "indirectly" supported by a source, which sends them in the wrong direction.
4. it is supported by a reliable source - that is not clear from the example. The example contains three statements followed by a mocked up citation: (1) Jones accused Smith of plagiarism; (2) Smith denied Jones' accusation of plagiarism; (3) Smith claimed that [X] is acceptable scholarly practice -- as well as a couple inferences: (4) Smith claims he was following acceptable scholarly practice; and (5) Smith's claim was made to refute Jones' accusation. So this is five things that must be verifiable and, if challenged, cited to one or more reliable sources. It is not that a passage of text needs to be supported, but that claims as to fact made in the passage need to be supported.
5. Now comes the... - too chatty and informal, almost a little condescending to the reader. Nothing "comes". Rather, keeping to the narrative voice about describing an editing process, it is more to the point to introduce the problem "But then..." or something like that.
6. unpublished synthesis - no, it is published. It is published on Wikipedia. Its status as synthesis does not depend on whether it is published or not. It is an incorrect to imply that synthesis is the combination in unpublished form of two published sources. We're going to explain in full detail why this new material is synthesis later on the page anyway. There's no point trying to telescope it here in the introduction, even if we could get it right.
7. ..of published material. Further on the above point, synthesis is the combination of two reliably sourced statements to generate a third that is not reliably sourced, not the combination of two published statements to derive an unpublished one.
8. The following material was... - unnecessarily points out the obvious, that the material that follows is following material.
9. that same Wikipedia article - redundancy / empty words: "that same" is no different than "the" or "that"; emphasis unnecessary. We've already established that the article is a Wikipedia article, no need to repeat the word Wikipedia.
10. just after... - "just" is an empty word, and imprecise to. In truth, the location does not matter. The whole phrase is unnecessary.
11. the above two sentences - more wordiness. We don't need to restate obvious irrelevant details (the sentences is above, there are two of them)

That explains why all of this is weak English. To fix it, it should be shorter, more to the point, avoid informality and narrative tone inconsistency, and avoid redundancy and the mention of obvious or impertinent details. To suggest a better version: "The above passage complies with this policy because all of the material is supported by a reliable source. However, an editor then added the following:" -- Wikidemon (talk) 23:12, 15 February 2009 (UTC)

1. I agree with your point.

2. That's an interesting point re "since" that I wasn't aware of. I looked in dictionaries and the timing definition is more prominent, although the logic definition is there too. As far as its appropriateness is concerned, it may be a matter of context. Where there are discussions involving reasoning and logic, it seems like "since" is most appropriate, from my recalled impression of the books and articles that I've read which involve reasoning. "Because" doesn't seem right, perhaps for the reason that you mentioned that it implies causation rather than deduction.

3. Perhaps you are correct that "directly" is not useful here. Although "directly" would be useful in other discussions in WP:NOR, here it is not needed and may only needlessly complicate the example.

4. I didn't understand this remark because we see that there is a citation and the statement said that the previous part of the example is supported by a reliable source. Why would anyone conclude that any statement in that part of the example is not supported by the citation? Perhaps it's the uncertainty because there isn't an actual source that goes with the citation? With all these considerations in mind, I think this aspect of the discussion is OK. It's not going to be perfect, but it may be best.

5. OK.

6. Re "no, it is published. It is published on Wikipedia." - I think it's clear that the discussion is not referring Wikipedia when discussing "published". However, there's another point here that is something that I've been considering for awhile, and might have been in your mind too while you were expressing your point. Wikipedia is full of proper synthesis. That's what an encyclopedia is supposed to do, bring together facts from various sources! Wikispeak seems to have perverted this definition of synthesis. Similar remarks can be made for the terms "original research" and "original thought". But those are other issues for other times to discuss.

Perhaps one of your ideas is that it is better to refer to the unpublished "conclusion" rather than the unpublished "synthesis", and I would tend to agree with you if that is one of the thoughts that you were trying to express. For example, better might be "the unpublished conclusion from the synthesis of published facts."

But perhaps your basic idea here is right, that it's better not to discuss it at all in the intro to the next passage of the example.

7. Actually, synthesis (in wikispeak) is the combination of two reliably sourced statements to form a conclusion that is not reliably sourced. As I somewhat mentioned above, almost all articles of the Wikipedia are combinations of reliable sources, and the combination (the wiki) isn't in a reliable source, unless the editor copied all the material in the wiki from a reliable source.

8. Hmmmmm, I'll pass on commenting on this one.

9. Agree.

10. Disagree. "Just" is necessary in the present wording to clear up any uncertainty in the reader's mind that there may have been intervening material that might have made it OK.

11. With my response to item 10 in mind, that phrase seems to be needed to go with "just".

But after all this, we seem to have ended up at the same place, i.e. I think your suggested edit is an improvement! For now, the only thing that I might suggest as a change in what you propose, is in the sentence, "However, an editor then added the following:". If it's the same editor that wrote the previous passage of the example, then I would change "an editor" to "the editor".

Perhaps for your next suggestion, you might show your proposed edit first and then make just a few comments about your main reason(s) for the edit. Then if there is an objection to something else in your proposal it can be discussed. But it may simply be viewed as OK without further discussion. Looking forward to it. --Bob K31416 (talk) 15:09, 16 February 2009 (UTC)

The examples

I think the example text should be cleaned up to more clearly illustrate the point.

The first paragraph:

Smith saysclaimed that Jones committed plagiarism by copying references from another book.another author's references.[cite] Jones denies this, and says it's responded that it is acceptable scholarly practice to use other people's books to find new references.[cite]

becomes:

Smith claimed that Jones committed plagiarism by copying another author's references.[cite] Jones responded that it is acceptable scholarly practice to use other people's books to find new references.[cite]

Reasoning:

1. The first paragraph is already greatly simplified from the original version regarding the Dershowitz–Finkelstein affair. If we're holding it out as a paragon of acceptable editing it should actually be done well without WP:MOS violations and WP:OR red flags.
2. "Says" and use of present tense is weak. This is an account of an unfolding dispute between two authors that happened years ago. Talking about historical texts in the present tense signals we're analyzing those texts -- just what this policy tells us not to do -- rather than covering biographical details about the authors or a notable dispute between them. Using past tense to cover an accusation and a response puts the focus on the authors, not their writings.
3. It is more of a claim than a statement. Although in general we avoid using colorful synonyms for "said", here it is important to note that Smith's writing was an accusation that Jones disputed.
4. "Smith says...Jones denies" - this rings synthesis alarm bells for me. One of the most common synthesis mistakes in all of Wikipedia, and a potent propaganda technique off wiki, is to take two different statements made at different times, and to sneakily claim that one is a response to or contradicts the other. To make up an example, "Hugo Chavez accused the United States of bullying by conducting a military training exercise off the coast of Venezuela. The United States denies this, claiming in a press release that it was a long planned operation necessary to maintain readiness." That would be fine if the press release were actually a response to the accusation by Chavez. But suppose the press release had nothing to do with the accusation by Chavez? An editor simply found two different statements, one of which seems to refute the other. To be ironclad, we should make clear, and source, that Jones' denial was in response to Smith's accusation. When I looked up the actual article's editing history, it seemed believable that this was Dershowitz' response to Finkelstin's accusation, but the sourcing was not ironclad. Without sourcing that statement B is a response to statement A, there is always a question that the paragraph is a synthesis of two unrelated facts. To smooth over this distraction, I've reworded it to claim explicitly that what Jones said is a response, not just a denial.
5. "references from another book" - for there to be another book there has to be a first book. There is no first book introduced here, only two authors. Better to say that Jones copied another author's references.
6. "denies this, and says it's" - three mistakes in five words! "This" is too informal and is an imprecise word to stand or an accusation of plagiarism. "and" makes it a run-on sentence. The sentence implies but does not come out and say that Jones' claim about acceptable sourcing is a denial - it uses the weasel-word "and" to urge the reader to make that connection. "It's" is too informal. Encyclopedic tone is to avoid contracted verbs.

- Wikidemon (talk) 21:03, 20 February 2009 (UTC)

I agree with much that you suggested. Here's a modification of your suggestion.

Smith claimed that Jones committed plagiarism by copying references from another author's book. Jones responded that it is acceptable scholarly practice to use other people's books to find new references.[cite]

1. I removed the citation of the first sentence so that both sentences come from the same source. Thus, there isn't the issue that you raised in your item 4.
2. In the first sentence, I used the phrase "references from another author's book", which corresponds with the phrase in the next sentence, "other people's books to find new references."

--Bob K31416 (talk) 21:48, 20 February 2009 (UTC)

Looks good to me. I've often found it better to separately cite within a paragraph multiple claims potentially in need of verification, even if to the same source, rather than using a single citation for the whole paragraph. That guards against the sloppiness where only part of the material is sourced and part is not, and also helps prevent the material from getting degraded if people subsequently add material from a new source to the middle of the paragraph. I won't belabor the point, because it's only marginally related to OR. Wikidemon (talk) 21:54, 20 February 2009 (UTC)

Synthesis example

As a general observation there are at least two common flavors of WP:SYNTH one sees on Wikipedia. The more common one in my experience is combining two verifiable facts to reach a conclusion, where the conclusion is a product of the Wikipedia editor's own thinking rather than something cited or verifiable to a source. Something like: Fact A[cite]. Fact B[cite]. Therefore, conclusion.[no cite]

The second type of SYNTH is combining a verifiable fact with a verifiable principle, then applying the principle to reach a conclusion. Something like: Fact A[cite]. Principle X[cite]. Therefore, conclusion.[no cite] The difference is a little subtle, but it leads to slightly different types of errors. But the point is that the first type mashes together two unrelated facts, and is fairly easy to spot. With the second type there is only one fact to work with, and people often get tripped over that because it is not as obvious that one should not analyze facts using logical principles. I don't know how common that is, but some editors around here refer to this second type of OR violation as original analysis, which is a sub-type of synthesis.

The example we have from the Dershowitz–Finkelstein affair is the second type, original analysis. It starts with a single fact, that Jones claims that it is acceptable scholarly practice to copy source lists. Then it adds a principle, that (according to the Harvard manual) it is an unacceptable practice but not plagiarism to fail to note the source consulted. Applying the principle to the facts at hand, it concludes that Jones violated Harvard's suggested practices. There are a few red herrings in this example, so it is not as clear as it could be. First, if the conclusion is right then both authors have made untrue statements - Smith incorrectly saying what Jones did is plagiarism, and Jones incorrectly saying that what Jones did is acceptable. Also, the paragraph in this example hints at this conclusion without saying it directly. I can't tangle it all out, but I think it would be clearer to the reader if we made the synthesis stick out in a more obvious way, without adding anything irrelevant.

I suggest changing the example from:

If Jones did not consult the original sources, this would be contrary to the practice recommended in the Harvard Writing with Sources manual, which requires citation of the source actually consulted. The Harvard manual does not call violating this rule "plagiarism". Instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them.

- to -

However, according to Writing with Sources[cite] all material of any sort copied from elsewhere requires a citation. Because Jones did not cite the book from which he obtained his source list, his practices violated the recommendation from that manual.

That is closer to the bare bones version of original analysis: Fact A + Principle X = Conclusion.

- Wikidemon (talk) 21:03, 20 February 2009 (UTC)

Note - looking through the encyclopedia, I see that I'm responsible for more than half of the references to WP:ANALYSIS, even though I did not create that redirect. So I'm an army of one trying to make a distinction between fact + fact, and fact + principle. Nevertheless, I do think the second paragraph in our example should be simplified and cleaned up, so it contains only a single, reasonably clear example of synthesis.Wikidemon (talk) 22:04, 20 February 2009 (UTC)

well, I get the distinction (though it is subtle). but really, isn't a principle just an implicit fact? (e.g., something that everyone knows is true, even though no one really says it outright). the problem on wikipedia is that people take two ideas that have broad consensus and combine them to produce a third idea that doesn't necessarily have the same consensus, but they present it as though it does (e.g. 'we know A' and 'we know B' therefore 'we know A+B', when in fact 'A+B' might not be something we know, depending on what the + means). --Ludwigs2 00:46, 21 February 2009 (UTC)

I think we need an opinion on that from the good folks at Wikiproject Epistemology. I don't want to get sidetracked, and probably the policy page shouldn't sidetrack the reader into a discussion of the theory of knowledge. But for what it's worth, I would say that all facts are things, but not all things are facts... except that there is a factual nature to all things, the fact of their existential conditions. Principles of logical deduction, rules, laws, and other abstract principles are a special sub-class of things. They are rules that act on other things - meta-things perhaps. They aren't really facts, except in the sense that there are facts about them. As an example, "In 1995 Bob grew a pumpkin that weighed 900 pounds" is a statement of fact that, on Wikipedia, would need among other things to be verified so that people know it is not original research. Consider a second fact, assumed to be properly cited: "In 1997 Cindy grew a pumpkin that weighed 950 pounds". It would by SYNTH to mash these two facts up to say "Therefore, Bob is not the world record holding pumpkin grower" and we could all imagine plenty of ways in which at type of OR could be misleading or just plain wrong. Maybe Bob grew a bigger pumpkin next year. Maybe Cindy's pumpkin was disqualified or uncertified. There is no proper basis for making any claims about world records, unless they are sourced. This other kind of synthesis I'm talking about, analysis, involves adding a deductive rule to the mix. For example, "The Guinness Book only began certifying pumpkins in 1996". It would be analysis to comment that "Bob's pumpkin was grown before the Guinness Book kept records, and is therefore not mentioned in the book". This second kind of synthesis is more seductive, and harder to deal with, because we do it everyday in our real lives. We're used to spotting factual inconsistencies and assumptions, but it's not so easy to spot leaps of logic. We don't know why Bob's pumpkin is omitted from the Guinness Book, not unless there is a source, ideally a third party source that says why. The Guinness Book often mentions unverified claims to records when it thinks they are credible and significant - they could have mentioned Bob's pumpkin if they wanted. It is not up to a Wikipedia editor to perform any but the most straightforward, uncontroversial analysis of facts. For example, "It snowed yesterday. It rained today. Because today comes after yesterday, the rain came after the snow." Nobody would question that level of analysis, and as someone commented elsewhere, to some extent all attempts to make an article out of multiple sources involves uncontroversial types of synthesis. The policy has to distinguish when the stringing together of different cited pieces of material crosses the threshold and becomes too speculative or untrustworthy. Wikidemon (talk) 21:29, 21 February 2009 (UTC)

Wikidemon, I think you're on to something but I also think you are losing the context of the first paragraph in your considerations. Here's the suggested first paragraph that we agreed on, along with your proposed second paragraph. Note that the issue is plagiarism.

Smith claimed that Jones committed plagiarism by copying references from another author's book. Jones responded that it is acceptable scholarly practice to use other people's books to find new references.[cite]

However, according to Writing with Sources[cite] all material of any sort copied from elsewhere requires a citation. Because Jones did not cite the book from which he obtained his source list, his practices violated the recommendation from that manual.

Would you consider modifying your suggested second paragraph, to incorporate the plagiarism issue, and fit the context of the first paragraph better? Thanks. --Bob K31416 (talk) 03:52, 21 February 2009 (UTC)

The following puts it more starkly: The style manual Writing with Sources recommends that material of any sort copied from elsewhere requires a citation, but does not specifically define copying references without attribution as an act of plagiarism. Therefore, Jones was incorrect in claiming his writing followed acceptable practices, but Smith was also incorrect in calling it plagiarism.

Here's another possibility:

Smith claimed that Jones committed plagiarism by copying references from another author's book. Jones responded that it is not plagiarism, but acceptable scholarly practice to copy references from other people's books.[cite]

Plagiarism is defined as using a source's information, ideas, words, or structure without citing them.[cite] Since Jones did not cite the book from which he obtained his references, but only listed the references themselves, he committed plagiarism.

--Bob K31416 (talk) 04:19, 22 February 2009 (UTC)

Good. I like that proposal best of all. It's not faithful to the original but I don't think that's a problem. That requires us to change the explanatory materials that follow so let's not implement it until we agree on that too. Wikidemon (talk) 04:35, 22 February 2009 (UTC)

Thanks!

Re "It's not faithful to the original ..." - Yes, I recognized that too. I think that all we have to do in that regard is change "based on" to "loosely based on" . --Bob K31416 (talk) 04:51, 22 February 2009 (UTC)

Regarding the explanation, it currently reads as follows:

The first paragraph is properly sourced, but the second paragraph is original research because it expresses the editor's opinion that, given the Harvard manual's definition of plagiarism, Jones did not commit it. To make the second paragraph consistent with this policy, a reliable source is needed that specifically comments on the Smith and Jones dispute and makes the same point about the Harvard manual and plagiarism. In other words, that precise analysis must have been published by a reliable source in relation to the topic before it can be published in Wikipedia by a contributor.

I would propose:

The first paragraph complies with this policy because each statement is properly sourced. However, the second paragraph fails because it introduces a new fact, a dictionary definition of plagiarism, and uses that fact to draw a conclusion not cited to any source, namely the Wikipedia editor's analysis that Jones' act fits the definition of plagiarism. Per this policy, that is considered a synthesis of the sourced facts, and is therefore original research. To make the second paragraph comply with the policy, a citation is required to a reliable source that specifically comments that Jones' work is plagiarism with reference to the dispute between Smith and Jones, in light of the cited dictionary definition. In other words, that precise analysis must have been published by a reliable source in relation to the topic before it can be published in Wikipedia by a contributor.

...except that it's two lines longer. Brevity is beautiful on policy pages... Wikidemon (talk) 05:35, 22 February 2009 (UTC)

This reads better. I would just like to suggest a slight change: and uses that fact to draw a conclusion not cited to any source, namely the Wikipedia editor's analysis that Jones' act fits that dictionary's definition of plagiarism. Cheers, Phenylalanine (talk) 20:01, 22 February 2009 (UTC)

Here's a try. It involves a little modification of the example which I have also included below.

---

Smith claimed that Jones committed plagiarism by copying references from another author's book. Jones responded that it is not plagiarism, but acceptable scholarly practice to copy references from other people's books.^[27]

Plagiarism is defined as using a source's information, ideas, words, or structure without citing them.^[28] Since Jones did not cite the book from which he obtained his references, but only listed the references themselves, he committed plagiarism.

Two parts of the above example are properly supported by references [27] and [28]. The part highlighted in bold uses information from each of the previous two parts, to form a conclusion that is in neither of the references. Furthermore, there is no other reference given for the conclusion. It appears that the conclusion was conceived by the editor, and is thus original research and not allowed.

---

--Bob K31416 (talk) 01:14, 23 February 2009 (UTC)

Does the following paragraph:

The "No original research" rule does not forbid routine calculations (e.g. adding or subtracting numbers, rounding them, calculating percentages, converting them into similar units, putting them on a graph, or calculating a person's age), or restatements, summaries, or rewordings, provided they are uncontroversial and add no new information to what is already present in the cited sources.

belong in the SYNTH section of this policy? In an edit, I suggested moving it to the "Sources" section, with some very slight rewriting to better relate it to the points contained within that section. It could also potentially be moved to the lede, although that would probably make the lede too long. It could also be moved to a new section altogether, although I think it fits tidily enough under "Sources".

However, I don't see how it relates to SYNTH at all. Does anyone else feel that "basic calculations are okay" has anything to do with "synthesizing published material to advance a position"?--Father Goose (talk) 00:35, 23 February 2009 (UTC)

Hi. I think that it does relate to SYNTH, since it has to do with, for example, taking a number from one source and adding it to a number from another source. The sum is in neither source and would thus violate SYNTH if that statement wasn't in the SYNTH section. --Bob K31416 (talk) 02:02, 23 February 2009 (UTC)

The main body of OR is adding pieces content based on the editor's personal knowledge (education, research, training, opinion, direct observation, etc) rather than a source. The other category, SYNTH, draws unsourced conclusions from verifiable information, either by (1) mashing together two or more pieces of information whose relationship to one another is established by the editor's own work rather than a reliable source; or (2) taking verifiable information and extending to reach a conclusion via the editor's own reasoning and analysis process rather than reporting a reliably sourced conclusion. The "simple arithmetic" is an exception to this second form of synthesis. It is okay to say that a circle of fabric was 4.0 inches (10.2 cm) in in radius,[cite] because that is not much of a stretch. However, it is original research to say that a circle of fabric was 4 inches in radius,[cite] and therefore it could not have been placed on a sphere of 12.0 inches circumference without forming a gap of at least 0.25 inches between the bottom of the fabric and the surface of the sphere, unless that conclusion is itself sourced with reference to the particular problem. A general knowledge of, or even a citation to, the mathematical principal involved, is not enough. That would be synthesizing one thing (the measurement) and another (the forumula) to produce an unsourced conclusion (the result). It makes some sense to group this with synthesis in general rather than the simpler basic case of reporting unsourced facts.Wikidemon (talk) 03:20, 23 February 2009 (UTC)

I ask beforehand for tolerance to my english.

There's a problem related to avoiding original research. It's inherent to the act of editing, no matter what one does to stay clear of it. Even in a mere sequencing of citations it's possible to perform a personal opinion.

I could notice, however, in some discussion pages (mainly in philosophy) users willing to expose their own opinions on the main articles' contents. What is, for instance, the cost of a kind of 'parallel' Wikipedia consecrated to these topics?

My suggestion is made under the observation that real knowledge neither is restricted to books, nor to their publisher's judgments - in fact the real source of knowledge is interchange, as explained by Protagoras and as we can witness nowadays. It could work as handwritten notes on the blanks of a printed page. Is this possible?--WReis (talk) 21:06, 23 February 2009 (UTC)

---

All this talk about what is and what is not published idea. I understand Wikipedia publishes ideas too. And I insist: all discussions in this topic are of the utmost importance, although the maximum approach to exemption of original research in whatever text anyone produces isn't perfection.

It is possible that allowing people to express their own ideas - in a separate section linked to the focused article - is the best way to exercise the search of attaining a higher level of impartiality in the main articles' expositions. Moreover, in doing so Wikipedia would be starting a really new way of composing encyclopedias, in which interchange of published material would be far more efficient than the traditional one, on paper.--WReis (talk) 09:56, 24 February 2009 (UTC)

---

I just found some good examples on what I come to say. They are in Omnipotence entry (other can be found in Omnipotence paradox's discussion page): one links to a pdf file by Prof. L. Pratt and two other to Internet pages on Judaism and Philosophy of Religion (this last one wasn't found). I didn't even read the documents to acknowledge their contents and I've never heard about their authors, nor they seem to be published out of the (supposed) authors' pdf and html formats.

Regarding the rules of Wikipidia's 'No original research', these files would be liable to be banned from the list of page's links. Am I right? Notwithstanding, personally, as well as conceptually, I don't see any reason to follow the rules in these as in several other cases. Regardless the way one could disagree to their contents, they undoubtedly enrich the encyclopedia.--WReis (talk) 11:28, 24 February 2009 (UTC)

If something can be verified by both primary and reliable secondary sources, should the latter be eschewed in favour of the former for simplicity or ease's sake? — pd_THOR ^{| =/\=} | 20:40, 25 February 2009 (UTC)

Recently, Father Goose raised some good points regarding two paragraphs that are now respectively in the Reliable sources and Synth sections. There may be a better place for them in their own section. Perhaps the proposed following section would be useful.

==Proofs and calculations==

If you are able to prove something that few or none currently believe, Wikipedia is not the place to première such a proof. Once a proof has been presented in a reliable source, however, it may be referenced. See: Wikipedia:Verifiability.

The "No original research" rule does not forbid routine calculations (e.g. adding or subtracting numbers, rounding them, calculating percentages, converting them into similar units, putting them on a graph, or calculating a person's age), or restatements, summaries, or rewordings, provided they are uncontroversial and add no new information to what is already present in the cited sources.

You may have noticed that the last part of the 2nd paragraph might be out of place here. Perhaps it should be incorporated in a sentence and placed in the lede.

The "No original research" rule does not forbid restatements, summaries, or rewordings, provided they are uncontroversial and add no new information to what is already present in the cited sources.

--Bob K31416 (talk) 14:25, 23 February 2009 (UTC)

I like it. Blueboar (talk) 14:36, 23 February 2009 (UTC)

Sorry. I added the last part before I saw Blueboar's message and after he posted it. --Bob K31416 (talk) 14:46, 23 February 2009 (UTC)

Actually, let me back off from my whole hearted approval... I like the intent and most of the lanugage. However, saying: "...it may be referenced", is inaccurate and could be seen as a conflict with WP:WEIGHT. Not all sources can or should be referenced. At the extreme end, I could see a Fringe theory fan pointing to the wording "... it may be referenced" to argue that he should be allowed to cite a fringe website that "proves" his favorite theory.

The point Bob is (correctly) trying to make is this: if you (a Wikipedia editor) came up with a proof, it is original research... if someone else came up with it (and published it), it isn't original research. So let's say exactly what we mean... May I suggest we change the phrase to "...it does not constitute original research." Blueboar (talk) 14:36, 23 February 2009 (UTC)

Please note that the two paragraphs already exist in the wiki. I did not create them. I am only proposing that they be moved to a new section. I think the issues you are raising are worth considering, but right now I am just proposing to move the paragraphs, and make the change mentioned in the last part of my message. Thanks.--Bob K31416 (talk) 15:00, 23 February 2009 (UTC)

Grouping these two together is still somewhat arbritrary. "Proofs", as used here, seems to refer to any kind of theory -- mathematical, scientific, or otherwise. I think it fits fine somewhere under "Sources", since it's a very basic type of original research -- "Aha! I have proved it!" -- to which we answer, "Don't prove it, source it."

Basic math is a general exception to original research, not really related to "I proved something". I perceive that the de facto rule on Wikipedia about basic math -- not yet reflected in our wording -- is that if a layman can check that it is correct, its use is okay. (This includes not just that the arithmetic is correct, but that how it is used is also correct: one can say that 12 + 3 = 15, but if the 12 is a quantity of cows and the 3 is a quantity of photons, the resulting quantity "15" is likely to be meaningless.)

Although some cases of math used in articles might involve numbers taken from one place and a formula taken from another, I don't see how that's necessarily a form of SYNTH, as argued by Wikidemon in the thread above. If the formula is either well-known or sourceable, and both the arithmetic and the aptness of its use are checkable by a layman, then it use ought to be okay; otherwise it's not. The "acceptable math issue" really deserves independent treatment from the "synth" issue. As I proposed with my earlier edit to NOR, I think it can fit in as the 2nd paragraph of == Sources == without too much trouble, by way of explaining that you don't need a source for basic math (though you might need a source to show that a given formula can be meaningfully applied to a given set of numbers).

The remaining issue, "restatements, summaries, or rewordings", is quite similar to something already in SYNTH: "Summarizing source material without changing its meaning is not synthesis; it is good editing." I think we could work "restatements and rewordings" into that paragraph as well.--Father Goose (talk) 22:32, 23 February 2009 (UTC)

Regarding your remark, "Grouping these two together is still somewhat arbritrary. "Proofs", as used here, seems to refer to any kind of theory -- mathematical, scientific, or otherwise." -

Actually, it's not arbitrary at all, as evidenced by your own remarks, viz. " 'Proofs', as used here, seems to refer to any kind of theory -- mathematical, scientific, or otherwise." Calculations and the other items, except for the items in the last part that begins with "restatements", are certainly mathematical. So a discussion of the acceptability of routine simple calculations would fit nicely with a discussion of unacceptable proofs, since those types of calculations are exceptions to the restrictions. --Bob K31416 (talk) 19:35, 24 February 2009 (UTC)

Arithmetic is not the only allowable synthesis tool. Simple statements of geometry (if A is north of B and B is north of C, A is north of C - but this does not necessarily work with east and west), astronomy (if the sun is out, it's daytime), metaphysics (if A and B did not overlap in time, then A followed B or B followed A), mechanical engineering (things may break into pieces but pieces do not break into things), etc., also fit that pattern. On the other hand obvious principles plus clever application, or dubious principles plus obvious application, need citing. The Pythagorean theorem is an example of the former. It is built on very simple premises, yet it would need to be cited because the proof, as brilliantly simple as it is, is not obvious to most people. Another example that frequently comes up is legal reasoning. On Wikipedia people often make the mistake of applying simple legal rules to reach an unsupported conclusion - for a current example, in the 1960s people born of a Kenyan parent were automatically British citizens and therefore Obama was a British citizen at birth. Or, doing X is against the law and person Y did X; therefore Y broke the law (a conclusion that could be wrong for dozens of reasons). Wikidemon (talk) 21:35, 24 February 2009 (UTC)

Bob K: Again, my point is that all mathematical proofs are examples of "I proved something", but not all "proved things" are mathematical proofs. The wording "If you are able to prove something that few or none currently believe, Wikipedia is not the place to première such a proof." is by no means restricted to mathematical proofs. "Don't prove stuff on Wikipedia" addresses a very basic form of OR, so it shouldn't be equated with "mathematical proofs", which are only a subset of the "don't prove stuff" dictum. Putting "don't prove stuff" under the heading "proofs and calculations" implies formal proofs, when we want to keep people from proving stuff like "OJ was the killer" as well.--Father Goose (talk) 02:41, 26 February 2009 (UTC)

I'm not quite sure that I understand all of your message. However, I'll try to respond to some of your message that I think I understand.

1) Re proofs - Would you care to suggest a better term?

2) Perhaps I might understand your remarks better if you quoted parts from my last message and addressed those points.

Thanks. --Bob K31416 (talk) 15:33, 26 February 2009 (UTC)

Wikidemon: I know that arithmetic is not the only allowable synthesis tool. My point is that arithmetic isn't necessarily a form of synthesis at all, in the WP:SYNTH sense. Multiplication algorithm contains quite a bit of math, for instance, but while the algorithms in it need to be sourced, the actual numbers and arithmetic passed through those algorithms don't. And such calculations are not "Synthesis of published material which advances a position" either -- so, again, we're back to my original point: calculations are not necessarily a form of synth. I suppose calculations can be involved in synth -- like saying that since Microsoft has sold 20 million XBox 360s, Playstation 20 million PS3s and Nintendo 25 million Wiis, there are 65 million owners of current-gen consoles out there. This would be false because plenty of gamers own more than one type of console. But the error there is still more mathematical in nature than a "synthesis of material to advance a position"; it's an adding of apples and oranges, which could easily be addressed with my photons-and-cows example, but is a stretch to explain via SYNTH. Synth is being pressed into describing all sorts of things that are certainly OR but not synth, and it just makes the SYNTH section more confusing than it already is. This is why I suggested moving the arithmetic-and-proof stuff out of it.--Father Goose (talk) 02:41, 26 February 2009 (UTC)

I've removed this [8] from the lead, as it has nothing to do with this policy, and the writing should ideally stay tight. Does anyone disagree? SlimVirgin ^{talk|contribs} 22:39, 26 February 2009 (UTC)

I agree, we don't need it here. I somehow doubt people will read it and say, "Oh! I'll stop bugging Wikipedia then."--Father Goose (talk) 23:14, 26 February 2009 (UTC)

Quite. :-) SlimVirgin ^{talk|contribs} 23:21, 26 February 2009 (UTC)

"This policy does not forbid routine calculations (e.g. adding or subtracting numbers) provided they add no new information."

One of the standard tenants of information science is that analysis of data is used to extract information (it is what data analysts get paid to do). So for example I may have a data set consisting of a series of dates of birth and a dates of death for a group of people. If I take that series and extract any additional information from it, I have new information. For example, I could calculate the age of the people in the data set. That is new information which I could then use as data to calculate the mean, median, mode, averages and other more sophisticated statistical analysis this would also be new information. Indeed once I have the age of the individuals, I could disregard the original data set as I now have a new data set to extract new information.

I understand what the sentence is trying to say but either "provided they add no new information" should be removed or it should be rephrased, because otherwise things like calculating age from birth and death dates is not valid according to the last phrase in the sentence. --PBS (talk) 10:52, 26 February 2009 (UTC)

I agree with you regarding the above statement. That statement was the result of a recent change here which has been reverted. Here is the actual statement that was, and is again in the wiki and notice that it doesn't have the problem that you were properly concerned with. It was very helpful that you noticed that.

The "No original research" rule does not forbid routine calculations (e.g. adding or subtracting numbers, rounding them, calculating percentages, converting them into similar units, putting them on a graph, or calculating a person's age), or restatements, summaries, or rewordings, provided they are uncontroversial and add no new information to what is already present in the cited sources.

--Bob K31416 (talk) 15:16, 26 February 2009 (UTC)

Bob, you reverted my attempt to tighten the writing. Can you say what the problem is? SlimVirgin ^{talk|contribs} 22:40, 26 February 2009 (UTC)

Small point here, but which sounds better "The best practice is", "Best practice is", or "It is best to"? I see all are in common use. Wikidemon (talk) 23:01, 26 February 2009 (UTC)

Best practice is "best practice." :-) Seriously, either that or "it is best to" would do. "The best practice" would be an odd construction, I think. SlimVirgin ^{talk|contribs} 23:23, 26 February 2009 (UTC)

SlimVirgin, You removed a significant amount of material without discussion. May I suggest in the future, if you want to do what you think is tightening, do it in smaller pieces, one at a time, and it will be much less likely to cause problems.

The "No original research" rule does not forbid routine calculations (e.g. adding or subtracting numbers, rounding them, calculating percentages, converting them into similar units, putting them on a graph, or calculating a person's age), or restatements, summaries, or rewordings, provided they are uncontroversial and add no new information to what is already present in the cited sources.

You removed the part in bold. This sentence was currently being discussed here. I think you and I had some similar thoughts about the material, but some differences too. So look at that section and see where we agree before you respond.

We differ on the part in parentheses in the above sentence. It is useful and shouldn't be deleted since it gives examples of other types of calculations that wouldn't be apparent from just the example of adding and subtracting.

--Bob K31416 (talk) 01:16, 27 February 2009 (UTC)

Can you say what was significant to the policy and was removed? The sentence in bold is overwritten e.g. restatements, rewordings, and summaries — what's the differece? Likewise, "routine calculations" covers all the examples you've given.

As you reverted the whole edit, not just this sentence, can you say what else you objected to and why? SlimVirgin ^{talk|contribs} 01:20, 27 February 2009 (UTC)

Please look at the other section that I gave a link for and respond to what I just wrote. Then I would like to discuss the other issues that you are concerned with. I think that the above issue is the most important one. Also, I noted that you reverted my revert. I haven't reverted it back, even though I feel it would be justified. I'm trying to have a productive discussion with you, and minimize what appears to be conflict, which may be more in your mind than mine. --Bob K31416 (talk) 01:40, 27 February 2009 (UTC)

I do notice that there is a part of your last message that we can discuss productively without your reading the other section first. Re "Likewise, 'routine calculations' covers all the examples you've given." - For tightening a regular article, you might have a point. But I think it would hurt this article on WP policy to remove those parts of the parentheses. As I mentioned, it gives examples of other types of calculations that wouldn't be apparent from just the example of adding and subtracting. If those other parts weren't there, I can envision a situation where, for example, someone puts numbers from a source or sources on a graph and someone would object that it violates WP:NOR. I don't think someone can point to a comment that just says "routine calculations" and "adding and subtacting" and expect anyone to believe that it means putting numbers on a graph. That's an example of why those other parts in the parentheses shouldn't be deleted. --Bob K31416 (talk) 03:14, 27 February 2009 (UTC)

Putting numbers on a graph might very well violate NOR. Depends how it's done. That's why leaving it at "routine calculations" is for the best. SlimVirgin ^{talk|contribs} 06:19, 27 February 2009 (UTC)

I'd like to suggest we get more explicit about "what basic math is okay" and what isn't:

This policy does not forbid routine calculations (adding or rounding numbers, putting them on a graph, converting units, calculating percentages or a person's age) provided editors agree that both the arithmetic and how it is used applied is correct. For instance, one can accurately say that 12 + 3 = 15, but if the "12" is a quantity of cows and the "3" is a quantity of photons, the resulting quantity "15" is likely to be meaningless.

Something like that. The "restatements, rewordings, etc." part belongs elsewhere, probably within the sentence "Summarizing source material without changing its meaning is not synthesis".--Father Goose (talk) 23:42, 26 February 2009 (UTC)

I don't understand the cows/photons point. SlimVirgin ^{talk|contribs} 06:19, 27 February 2009 (UTC)

SlimVirgin, It appears that you now believe that, regarding the rest of the part in the parenthesis that was deleted, it is not a matter of just being superfluous, but you believe that at least one part of it may be allowing something that is not consistent with NOR. Is that your position now? --Bob K31416 (talk) 12:57, 27 February 2009 (UTC)

Upon further reflection, the cows/photons thing is probably superfluous. A more realistic example of "misapplied arithmetic" would be trying to add together percentages of things not belonging to the same set. But it's probably best to leave "how it is applied" undefined rather than back ourselves into a corner, as I believe you have been arguing. But I think the first sentence by itself might be worth considering (I tweaked "how it is used" to "how it is applied").--Father Goose (talk) 20:02, 27 February 2009 (UTC)

I would like to propose WP:NORDR as a supplemental essay which would be linked in the "see also" section of the policy. I think the essay provides some useful details about the policy with specific examples to illustrate the principles outlined. Thanks in advance for your consideration. --Phenylalanine (talk) 04:40, 27 February 2009 (UTC)

The difficulty with linking to essays is that they might contradict the policy, or someone might add a contradiction, so they would have to be watched closely. SlimVirgin ^{talk|contribs} 06:21, 27 February 2009 (UTC)

Yes, I disagree with "officially" linking essays in all but the most necessary cases. For example, WP:SILENCE as part of WP:CONSENSUS. Not to mention that WP:NORDR is a bit in contradiction what in generally has been understood is the spirit of NOR and certainly a bit WP:CREEPy (ie it is a very long essay).--Cerejota (talk) 23:44, 27 February 2009 (UTC)

As I prototyped here, I am proposing a general restructuring and streamlining of WP:SYNTH. I've always found this section muddled and unclear, but there's no reason it should be. I'll step through each of the changes I made and why; I'd appreciate people weighing in with agreements and disagreements.

Here's the current version of the first paragraph:

“

Synthesis occurs when an editor puts together multiple sources to reach a conclusion that is not in any of the sources. Even if published by reliable sources, material must not be connected together in such a way that it constitutes original research. If the sources cited do not explicitly reach the same conclusion, or if the sources cited are not directly related to the article subject, then the editor is engaged in original research.

”

First sentence

"Synthesis occurs when an editor puts together multiple sources to reach a conclusion that is not in any of the sources."

• Ugh. We're in bad shape if we're opening this thing with the passive voice. (See Elements of Style.) What we want to say is "You, editor: do not do this", then explain in greater detail what "this" is and why it's bad. Here's what that sentence could look like in the active voice:

Do not put together multiple sources to reach a conclusion that is not in any of the sources.

Second sentence

We can drop the second sentence altogether and lose nothing:

• "Even if published by reliable sources," - We don't need to mention "reliable sources" here if synthesizing facts from any kind of source is a no-no. Dragging RS into this is a distraction, maybe even WP:BEANS.
• "material must not be connected together" - We said that in the first sentence.
• "in such a way that it constitutes original research." - We don't need to point out that SYNTH is OR; that's amply explained by the fact that SYNTH is a section in NOR.

Third sentence

• "If the sources cited do not explicitly reach the same conclusion," - ...the same conclusion as each other? There's an ambiguity in the language here.
• " or if the sources cited are not directly related to the article subject," - direct relation is an important concept here, but I think we can better address it using different language. I'll get to that in the next section.
• "then the editor is engaged in original research." - Again, we don't need to point out that SYNTH is OR. It's like we're pleading to ourselves that this is true.

And, rewritten

So. Let's simplify our explanation of SYNTH by just saying "Hey! Don't do this; here is why."

Rewritten as I propose:

Do not invoke a source to back up an assertion not originally made in the source. This includes connecting multiple sources together to express a conclusion not explicitly stated in any of the original sources.

This outlines the general principle of how to use (and not use) sources, then mentions how SYNTH is a violation of that general principle: if the sources didn't reach an exact conclusion, ("Jones committed plagiarism") you shouldn't invoke them to reach that conclusion either.

In the rewritten version, I didn't include the words "not directly related to the article subject", as I think this idea can be covered more broadly by the words "[do not] express a conclusion not explicitly stated in any of the original sources". If the sources don't make the exact same assertion that you do ("Jones committed plagiarism"), they are "not related to the subject" at hand. Further, I think we're better off not using the words "the article subject" anyhow; a source can be completely germane to a small but important point in an article without having any relation to the article's general subject.

From here, we should jump right into the example, Smith v. Jones, to immediately drill in what we're talking about, and why it's bad. The second part of the current paragraph, "Summarizing source material... is not synthesis" is better saved as a clarification that comes after the example.

That's the first set of changes I'd like to discuss. Comments?--Father Goose (talk) 23:10, 26 February 2009 (UTC)

It looks pretty good, except that the new first sentence is confusing. Written abstractly as it is one has to think a while to figure out what kind of edits it is implicating. I think the problem is that "back up" is not specific enough. Perhaps a source could be validly used to clarify, expand on, give an example of, or do something else that you could call "backing up". How about "reach a conclusion", "support a conclusion", "verify a statement", or something more to the point if that's what is meant? Wikidemon (talk) 23:46, 26 February 2009 (UTC)

Support's good. With that first sentence, I'm not trying to "implicate" any specific edits, but to restate the general principle of "stick to the sources", namely, "don't say something other than what the sources say". In the second sentence, I outline how synthesis is a specific form of this error: "don't string sources together to reach a conclusion other than what the sources conclude". Given that, I guess I should also change "express a conclusion" in the second sentence to "support a conclusion":

Do not invoke a source to support an assertion not originally made in the source. This includes connecting multiple sources together to support a conclusion not explicitly stated in any of the original sources.

--Father Goose (talk) 04:35, 27 February 2009 (UTC)

I cut and pasted together the three sentences of Father Goose's (FG) suggested revision.

Do not put together multiple sources to reach a conclusion that is not in any of the sources. Do not invoke a source to back up an assertion not originally made in the source. This includes connecting multiple sources together to express a conclusion not explicitly stated in any of the original sources.

And here's what's in the wiki now for comparison.

Synthesis occurs when an editor puts together multiple sources to reach a conclusion that is not in any of the sources. Even if published by reliable sources, material must not be connected together in such a way that it constitutes original research. If the sources cited do not explicitly reach the same conclusion, or are not directly related to the article subject, the editor is engaged in original research.

I have a problem with the first sentence of the wiki too, because I believe it is false. Synthesis is entirely proper and is the way wikis are properly put together. Reaching an unsourced conclusion by synthesis is the activity that is improper. So we shouldn't use the term synthesis to characterize an improper activity and I note that FG's version correctly doesn't do that. The second sentence of FG's suggestion doesn't seem to be about combining information from multiple sources so it seems out of place. The third sentence seems redundant with the first.

So, I agree with some of Father Goose's suggestion but I would make the following modifications which are indicated below with strikeouts and bold.

Do not put together information from multiple sources to reach a conclusion that is not explicitly in any one of the sources. Do not invoke a source to back up an assertion not originally made in the source. This includes connecting multiple sources together to express a conclusion not explicitly stated in any of the original sources.

Here it is without changes noted.

Do not put together information from multiple sources to reach a conclusion that is not explicitly in any one of the sources.

--Bob K31416 (talk) 17:38, 27 February 2009 (UTC)

So that would be in place of the two sentences I suggested, or the three that are currently in the policy? If so, I give a thumbs up to your single sentence -- though I'd tweak it to:

Do not put together information from multiple sources to support a conclusion that is not explicitly stated in any of the sources.

--Father Goose (talk) 20:25, 27 February 2009 (UTC)

Tweak OK. For now, how about just replacing the first sentence, and then we can all look at the rest of the paragraph with the first sentence in place. Also, I think that making changes in smaller steps, when possible, is more considerate of other editors and leads to better editing. I think this is a case where we can make just this first step for now. It's OK with me if you replace just the first sentence with the above sentence that we agreed on. --Bob K31416 (talk) 21:02, 27 February 2009 (UTC)

I know what it's getting at, but does 'explicit' mean exactly what we're using it for here? Per Wiktionary 'explicit' refers to a degree or threshold of specificity, detail, or clarity.[9] I think we're looking for a level of definiteness, i.e. there for sure, for real, not there by implication.Wikidemon (talk) 21:09, 27 February 2009 (UTC)

Here's the first definition of explicit from the Merriam-Webster dictionary.[10]

1 a: fully revealed or expressed without vagueness, implication, or ambiguity : leaving no question as to meaning or intent

According to this, "explicit" looks like the right word. The Wictionary is not made by professional experts, like the Merriam-Webster is, as far as I know. --Bob K31416 (talk) 21:21, 27 February 2009 (UTC)

I don't think the word quite captures what we're trying to say. It seems to be there for emphasis more than for clarity. If I had the right word I would propose it. The problem is that 'expicit' comes with some baggage having to do with the level of detail and demonstrative assertiveness. Consider "I described each and every step in explicit detail" or "I explicitly told her not to put that in the fridge". Per our policy we are allowed to rephrase, summarize, restate in our own words, etc. Therefore, an allowable conclusion may be in the source in a way that is not particularly explicit. We can cite a source saying "it snowed that day" to stand for the proposition that "there was precipitation" that day, even though the source does not explicitly say there was precipitation. Or "X was older than Y and Z" can support "Z was younger than X" even though it does not explicitly say so. A better word, if it even needs a word for emphasis, might convey that the conclusion is in the source without any doubt, incontrovertibly so, without any further assumptions, leaps of logic, etc. Wikidemon (talk) 19:02, 28 February 2009 (UTC)

...I'll add that this is not a strenuous objection, more like a word choice thing. I don't think it's deadly, and if no better word is found it's probably better to use "explicitly" for emphasis than to not emphasize the point. - Wikidemon (talk) 19:07, 28 February 2009 (UTC)

<outdent> Wikidemon, After reading your message, and giving it more thought, it doesn't appear that "explicitly" is needed. Here's what it would look like without "explicitly".

Do not put together information from multiple sources to support a conclusion that is not stated in any of the sources.

It looks like the word "stated" is sufficient by itself. With "explicitly" removed, would you agree to just the replacement of the first sentence in the section for now, with the option of altering the rest of the first paragraph later? --Bob K31416 (talk) 19:42, 28 February 2009 (UTC)

Sure. Wikidemon (talk) 20:13, 28 February 2009 (UTC)

Great, just a minor change:

Do not put together information from multiple sources to arrive at a conclusion that is not stated in any of the sources.

Strictly speaking, it is the sources that serve to "support" the conclusion, not the process of putting together information. Cheers, --Phenylalanine (talk) 00:05, 1 March 2009 (UTC)

The first version of the change used "reach". The second version used "support". And now a third suggestion is to use "arrive at", which I think is a synonym for "reach". I have no objections to any of these. Since this seems to be a fine point that wouldn't significantly affect the sentence one way or the other, and since the version with "support" is just about ready to go into the article, how about deferring your suggestion until after the sentence is in the article. In fact, if the sentence was put in the article, and you immediately changed "support" to "arrived at", no one may even object. I sure wouldn't. So just to move things along, I'm going to put the sentence in the article now. I hope no one's offended. --Bob K31416 (talk) 01:08, 1 March 2009 (UTC)

Reach, support, arrive at -- I think they'll all work, it's just a question of word choice at this point.

I've also gone ahead and removed the two sentences that follow it, as they are now thoroughly redundant with our new first sentence and the paragraph that follows it. If anybody thinks those sentences should be retained, I suppose we will soon hear from you.--Father Goose (talk) 21:27, 2 March 2009 (UTC)

Yesterday a section on translations was added that was based on a previous discussion on this talk page. The subject of translations is already discussed elsewhere in the Wikipedia at WP:NONENG which is part of the policy WP:V. I agree that it is useful to mention translations at WP:NOR, but in an abridged form, since it is already discussed at WP:NONENG. I would suggest changing the recent addition to the following.

==Exceptions==

• Using non-english sources - Wikipedia editors may supply a translation of material from a non-english source, or use material from a non-english source, only if it is done in a way that is consistent with the policy described at WP:NONENG.

--Bob K31416 (talk) 15:24, 1 March 2009 (UTC)

Hmmm... I kind of liked the longer language. WP:NONENG really focusses on the source (and says to provide a translation to make verifiability easier) while the passage that was recently added here focuses more on the translation (and how to avoid OR when doing so). The two policy statements are definitely related(and should be linked to each other), but I think they are talking about different parts of the process. Blueboar (talk) 16:01, 1 March 2009 (UTC)

---

For comparison, here's WP:NONENG.

=== Non-English sources ===

Because this is the English Wikipedia, for the convenience of our readers, editors should use English-language sources in preference to sources in other languages, assuming the availability of an English-language source of equal quality, so that readers can easily verify that the source material has been used correctly. Where editors translate any direct quote, they should quote the relevant portion of the original, non-English text in a footnote or in the article. Translations published by reliable sources are preferred over translations made by Wikipedia editors.

---

And here's what was recently put into WP:NOR.

===Translations===

Where English translations of non-English material are unavailable, Wikipedia editors proficient in the foreign language involved may supply a translation citing Wikipedia editors as the source of the translation. Such translations may be challenged on the discussion page of articles involved. If such translations are challenged, editors should cooperate in producing a consensus-supported translation. Copyright restrictions permitting, translations published by reliable sources are preferred over translations provided by Wikipedia editors.

---

And before the new section was added to WP:NOR it was discussed here .

Blueboar, I don't understand your comment, " WP:NONENG really focusses on the source (and says to provide a translation to make verifiability easier) while the passage that was recently added here focuses more on the translation (and how to avoid OR when doing so)."

It looks like what the new section has that WP:NONENG doesn't have is comments about challenging a translation, working towards a consensus, and a copyright issue. This seems like just the general process of challenging any edit and modifying it to form a consensus, whether it involves a translation or anything else. Similarly it seems that the issue about copyright is covered by Wikipedia policies about copyright in general. Anyhow, if either the challenging aspect or the copyright aspect, are worth discussing in a policy on translations, there doesn't seem to be any reason why they can't be discussed in the established policy at WP:NONENG, if they are worthwhile.

WP:NONENG states,

"Where editors translate any direct quote, they should quote the relevant portion of the original, non-English text in a footnote or in the article."

And here's the corresponding part of the new addition to WP:NOR,

"...Wikipedia editors proficient in the foreign language involved may supply a translation citing Wikipedia editors as the source of the translation."

The policy regarding translating quotes seems to be explained at least as well at WP:NONENG. Furthermore, the policy at WP:NONENG goes beyond translations of quotes since it mentions any use of non-english sources when it states,

"Because this is the English Wikipedia, for the convenience of our readers, editors should use English-language sources in preference to sources in other languages, assuming the availability of an English-language source of equal quality, so that readers can easily verify that the source material has been used correctly."

So I would suggest that the version at the top of this Talk section, or something like it, should be used instead of the recent addition regarding translations. Thank you. --Bob K31416 (talk) 00:38, 2 March 2009 (UTC)

Synthesis is the only way an encyclopaedia can be written. Sometimes facts do not have to be published to be scientific. This policy simply encourages plagiarism. What a load of BS. This policy is what turned wikipedia from an amazing place for the collective sum of all human knowledge into just another mainstream encyclopaedic work, but one which is prone to POV and weasle language, by those people "reporting" (plagiarising but escaping that definition by simply re-phrasing) what has been ublished already. No synthesis policy prevents wiki from achieving the goal we are to imagine, "the collective sum of Human knowledge". Rather wiki becomes the collective sum of idiotic rephrasal of already published work by people who don't know any better because their creativity and knowledge is limited to regurgitation. BS, BS, BS, BS, :) I hoppe I found the right place to post my insignificant opinion. Thank you for ignoring me.82.6.30.147 (talk) 10:36, 24 February 2009 (UTC)

Re "BS, BS, BS, BS, :) I hoppe I found the right place to post my insignificant opinion. Thank you for ignoring me." -

A bit of the Uncyclopedia has invaded. Thanks for the endorphins. --Bob K31416 (talk) 14:28, 24 February 2009 (UTC)

Actually, I can give you a more serious answer. Regarding your remark, "Synthesis is the only way an encyclopaedia can be written." I made some comments like this previously myself. "Wikipedia is full of proper synthesis. That's what an encyclopedia is supposed to do, bring together facts from various sources! Wikispeak seems to have perverted this definition of synthesis. Similar remarks can be made for the terms 'original research' and 'original thought'." An editor does original research every time the editor goes looking through the literature for material and uses original thought in combining and expressing them in the Wikipedia. --Bob K31416 (talk) 21:42, 24 February 2009 (UTC)

82.6.30.147: a policy is what people make out of it. some people have a very narrow view about original research, while some people have a broader view; some people use this policy well, while some abuse it. I understand your objection, but you (I'm sure) can think of any number of wild theories and speculations that have no place in an encyclopedia (regardless of how firmly convinced the people who advance them are). --Ludwigs2 00:40, 25 February 2009 (UTC)

A US Supreme Court justice once made a comment about judging whether laws are valid or are instead too vague to stand, which I think applies here. It is that rules should not be judged based on whether they permit bad results under extreme interpretations, but rather whether in the real world sensible people interpret the rule in a way that is useful and was intended by the drafter. Something like that. Wikipedia is full of such rules. I think the key comes from WP:V. Things need to be verify-able, not necessarily verified. OR applies in a similar way. Observations and conclusions subject to reasonable questioning need to be cited to a reliable source. Certain simple things that everybody knows, and have next to zero chance of being wrong, don't.Wikidemon (talk) 01:00, 25 February 2009 (UTC)

Since last feb. 23 I started a sort of campaign to search in community a solution to the natural wave of human thinking watering the pages of Wikipedia. I believe, like the majority, precision might be a continual goal in spite its perfect acquisition isn't but an unattainable desire. So, the tendency of thought is to continually produce changes in whatever it makes contact with, a stream we can't oppose with a handful of rules. Aside mainstream thinking, anyone feels oneself able to emit one's original view on any subject. Considering the difficulties involved in exposing one's ideas through mainstream, imposed notably by the 'academy', whose democracy works (if so) just among its members, I ask: why not state in Wikipedia a free zone of thought, each of its pages or topics linking to an article it refers, and in which everyone could add comments, make one's original researches and even contribute to the refinement of the main articles? In a certain sense this is already being done when people link pdf papers and html documents to the articles' pages in Wikipedia, comment in discussion-related pages (I give some examples above), and even here, in NOR section.

I'm sure 'academicians' would be the most frequent users of this new section, where undoubtedly they will renew their dusty ideas. For caution it would be advisable that the page of each contributor couldn't be edited except by the owner.--WReis (talk) 11:05, 25 February 2009 (UTC)

As far as I know, WP:NOR doesn't apply to an article's talk page. So those discussions may already be allowed on talk pages. Does anyone know anything more about this? Thanks. --Bob K31416 (talk) 15:17, 25 February 2009 (UTC)

It sounds like wikinfo is what you're looking for. The article talk pages are intended solely for discussions related to the development of the article, and personal opinions of the topic's subject are discouraged, and in highly controversial articles' talk pages such commentary is frequently removed on sight. (See "Talk pages are not a forum for editors to argue their own different points of view about controversial issues", Talk page guidelines ). Professor marginalia (talk) 15:33, 25 February 2009 (UTC)

As English is not my native language, I certainly didn't make my thoughts clear. Let me try again.

I believe I understood that this talk page for NOR is used to deal with more general aspects of the composition of articles, and that talk pages created for normal articles might contain specific observations on the articles they refer. If my beliefs are correct, these points are out of doubt to me.

To state the problem let's start, for instance, with (1) Talk:Omnipotence paradox. I ask you to take a general look at the entries in this page: not only one of them is commenting the content of the article, say, there are several of them exposing ideas or opinions of their authors about the way philosophers and theologians enrolled in the article used to deal with this paradox. (2) In a related subject, Omnipotence, I could find among the links (at the bottom of the page) one pdf and two html documents (the second one 'not found') with personal researches on omnipotence, not exactly clearing some point of the classic authors cited. I believe I witnessed in these Wikipedia pages - and in several other I used to visit years ago - a stream we can't control with some rules, no matter how complex or strict they could be. People want to expose what they are thinking, but have no better places to do it except these corners inside Wikipedia.

I don't think this attitude perverts the essence of any true encyclopedia. This is the opposite: it enriches it. So, what about creating a specific place for this practice inside Wikipedia? How much in gigabites - or money - would it cost? I see some advantages in implementing the idea, for I believe there is a lot of good thinkers whose participation in mainstream circles of knowledge is not allowed, as we can see in the mentioned pages.

Yes, I admit, it could be as simple as to add a list of links to existing pages, but I imagine this proposal could work with Wikipedia users' pages: the OR contribution on specific subjects in each user page would be listed in a links' section in the related articles pages.

I just want to know what would Wikipedia community think about the idea. I admit it could make a 180^o turn in the conception of encyclopedias, although some sites already collect general information from people otherwise anonymous, like life/family histories, images etc. I fear I'm thinking after Borges' 'Babel's Library', trying to apply the concept to Wikipedia. Perhaps it's a dangerous idea.--WReis (talk) 19:11, 25 February 2009 (UTC)

Because what you are proposing is not an encyclopedia, but the creation of a think tank. -- The Red Pen of Doom 19:14, 25 February 2009 (UTC)

Yes, TheRedPenOfDoom, you're quite right. I truly think encyclopedias are - or should be - think tanks, specially when using tools like bits and bytes (on paper this would be effectively impossible), but I understand this is the desire of Wikipedia community: to keep the Wikipedia as close as possible to mainstream encyclopedias. This is just a matter of choice.

Thanks to my foolishness I didn't try the link to Wikinfo, wich Professor marginalia so gently posted above, before spending so much time (yours and mine) writing the last post: the site is exactly what I'm looking for. Thank you.--WReis (talk) 20:05, 25 February 2009 (UTC)

I guess that I will just disagree with your premise. What you are proposing sounds a lot like UseNet forums and no one ever accused UseNet forums of turning out buckets of encyclopedic material. -- The Red Pen of Doom 20:51, 25 February 2009 (UTC)

TheRedPenOfDoom expresses the principle well. I'd go one step farther. WP:NOR as a basic principle is not unique to Wikipedia. No product called an encyclopedia should properly be doing original research nor original synthesis. Perhaps the anon-IP contributor who posed the question is failing to understand the distinction between doing original synthesis of concepts on the one hand, differentiated from writing unique encyclopedia articles on the other hand. It's an important difference, and this policy serves, in part, to help WP users identify the difference between the two. ... Kenosis (talk) 02:29, 2 March 2009 (UTC)

excert from another page:

The original research policy tends to break the system. For example, it says to an archeologist; "Yes, you did discover the lost kingdom of Atlantis, but you may not post that on Wikipedia. Only if the New York Times makes writes an article on the subject can the information be written to Wikipedia."

Completely unfair and unproductive.

Spitfire19 (Talk) 19:59, 1 March 2009 (UTC)

Wikipedia doesn't do original reserach because there's no way to verify your claims. Who then was a gentleman? (talk) 22:55, 1 March 2009 (UTC)

So the question is, what, in the real world, makes something true?Spitfire19 (Talk) 00:40, 2 March 2009 (UTC)

Actually, if an archeologist did discover the lost kingdom of Atlantis, the first thing he would probably do is notify the news media... if he were at all a professional, the second thing he would do would be to write an article about his discovery for a peer reviewed journal. Then he might write a book about his discovery, or sign a TV contract with National Geographic or the Discovery Channel so he could make some money off of it. Posting his discovery on Wikipedia would probably come fairly low down on his list of priorities. Blueboar (talk) 16:30, 2 March 2009 (UTC)

(edit conflict) In the real world, people don't agree.[11][12] Regardless, Wikipedia has principles that have much more in common with information science and knowledge management, than with the pursuit of truth or epistomological inquiry. The latter matters are for an entirely different project model and scope. Vassyana (talk) 16:46, 2 March 2009 (UTC)

Regarding Spitfire19's question, "So the question is, what, in the real world, makes something true?" - Consensus ? --Bob K31416 (talk) 18:54, 2 March 2009 (UTC)

We adopted the NOR policy, an adjunct of our verifiability policy, because we don't know whether a person claiming to be an archeologist is actually one. Further, even if we can confirm that someone posting to Wikipedia is an archeologist, we still can't confirm if their claim of having found Atlantis is true. Nor can we rely on individual Wikipedians to research the claim (in the manner of a journalist), because we don't know how methodical those individuals are.

However, we can have a reasonable degree of confidence that publications with editorial oversight -- newspapers, (non-self-published) books, peer-reviewed journals -- are not written by crackpots or vandals. Sometimes such publications contradict one another: our publication can still remain factual by not embracing one view or another, but by noting the existence of such contradictory views. Limiting ourselves to describing only what can be found in reliable sources is key to maintaining Wikipedia's reliability.--Father Goose (talk) 20:50, 2 March 2009 (UTC)

If we just listed people's claims as truth then the Bigfoot article would state that it was proven real by those hunters who "found" it last year. --Bill ^{(talk|contribs)} 20:41, 3 March 2009 (UTC)

There's a problem with the example as currently listed. It includes the statement:

Plagiarism is defined as using a source's information, ideas, words, or structure without citing them.

However, this insertion is itself a version of Original research. Why do we insert a specific definition of plagiarism from a specific source at this point? Is it not to weight in some way the argument listed immediately prior? And if you say, "this is merely a definition", then why not simply link the word plagiarism? Why instead choose that definition from that source? Jayjg ^(talk) 02:41, 3 March 2009 (UTC)

I don't know what you're getting at. Choosing one version of sourced content and one particular source, over another version and source, is a matter for many other policies and guidelines, and editorial discretion. Here the issue presupposes that the editor has made the choice and the sourcing is adequate. We use a specific, arbitrary, example to illustrate the problem with synthesis and why it constitutes original research. We recently streamlined the example so as to be clearer and not include red herrings. In the example, the definition of plagiarism is cited to hypothetical source #28, so that by premise it is not original research. The definition of plagiarism here is arbitrary, as is everything else in the example. It is an example. A wikilink in the example would be distracting, and beside the point. Wikidemon (talk) 02:55, 3 March 2009 (UTC)

You may be hitting on a subtle (IMO) distinction we talked about a couple weeks ago between synthesis being a matter of mashing cited fact + cited definition / principle to form uncited conclusion (which I called "analysis"), and synthesis being a matter of mashing cited fact + cited fact (via uncited / obvious rules of deduction) to form uncited conclusion (which I see as a more garden variety of synthesis). The former is what we have in the example: #1 A did X [CITE]; #2 X is defined as Y [CITE]; #3 therefore, A did Y [UNCITED SYNTHESIS]. However you look at it, the example does illustrate a common problem that this policy is trying to prevent.Wikidemon (talk) 03:04, 3 March 2009 (UTC)

The issue I'm raising is that there is no reason to include a specific definition of "plagiarism" in the article about the original controversy - Wikipedia already has an article on the topic, so all that needs be done is provide a link to plagiarism. Including that specific definition and citation is itself a form of original research. Leave the definitions to other Wikipedia articles, or, to an article that defines "plagiarism" specifically in the context of the dispute being discussed in the article. Jayjg ^(talk) 03:53, 3 March 2009 (UTC)

If someone wants to state in an article what plagiarism is they need a cite in that article to a source that supports the statement, not just a link to the Wikipedia article on the subject. That would be a sourcing problem in itself. By contrast, it is not OR to state in any article what plagiarism is, cited to a source. Which source/definition to use is a decision by an editor to be sure, but I don't see how it is OR. OR is a claim that comes without a source. Here there is a source. Do you think it would be clearer if instead of calling it a "definition" we posed it as something else? For example, instead of Plagiarism is defined as using a source's information, ideas, words, or structure without citing them[28] we could say When an author copies a source's information without citing the source, that is one form of plagiarism.[28] Wikidemon (talk) 04:05, 3 March 2009 (UTC)

How do you know that that particular definition of plagiarism is relevant to the issue being raised in the article - i.e., the controversy being described? Jayjg ^(talk) 04:17, 3 March 2009 (UTC)

Jayjg, Like Wikidemon wrote, "I don't know what you're getting at." I can say for myself, I don't know what you're getting at either!

You wrote, "How do you know that that particular definition of plagiarism is relevant to the issue being raised in the article - i.e., the controversy being described?"

You seem to be working on some article about a dispute, not the WP:NOR policy page. --Bob K31416 (talk) 05:02, 3 March 2009 (UTC)

...so to answer Jayjg's question literally, we don't know that this particular definition of plagiarism is relevant to the issue at hand. It's just an example. We're assuming that the editor who wrote it chose a good definition. But the fact that this particular definition is not really germane to the issue, and that the piece of prose contains original research, is precisely the point of the example. We're showing by counter-example that when you choose one fact to describe, and one definition to provide, it is original research to apply one to the other. Wikidemon (talk) 05:16, 3 March 2009 (UTC)

Ah, but the editor shouldn't be choosing any definitions at all. As stated, the editor can rely on the link to described what plagiarism is (for those who are unsure). And if a definition must be provided, then, as with all other Original Research, we should be relying on what Reliable Secondary sources say is a relevant definition of plagiarism. That is, if the first sentence on the subject is:

Smith claimed that Jones committed plagiarism by copying references from another author's book. Jones responded that it is not plagiarism, but acceptable scholarly practice to copy references from other people's books.

then a permissible second sentence would be:

According to a New York Times article on the controversy, plagiarism in this case is defined as "xyz".[28]

Without a secondary source giving a relevant definition in this specific case, insertion of any particular definition is Original Research. Jayjg ^(talk) 05:29, 3 March 2009 (UTC)

And, by the way, that is why the original formulation of the example was more accurate. It stated:

The second paragraph was original research because it expressed the editor's opinion that, given the Harvard manual's definition of plagiarism, Jones did not commit it. To make the second paragraph consistent with this policy, a reliable source would be needed that specifically comments on the Smith and Jones dispute and makes the same point about the Harvard manual and plagiarism. In other words, that precise analysis must have been published by a reliable source in relation to the topic before it can be published in Wikipedia by a contributor.

And that is exactly right; it isn't just the fact that the conclusion was not reached by the source, but also the fact that the Harvard manual was brought as relevant to the specific case at hand. There's no indication that the Harvard manual is indeed relevant to the case at hand, unless a reliable secondary source says it is. Jayjg ^(talk) 05:42, 3 March 2009 (UTC)

(reset indent) Jayjg is quite right here; "plagiarism is defined as x, y, z" is a problematic statement unto itself, as "what is plagiarism" will differ depending on whom you ask. (Jones, for instance, defines it differently from Harvard -- and maybe Jones is as much of an authority on plagiarism as the Harvard authors are.)

We would be making the same error if we wrote the "Smith claims Jones plagiarized" sentence as "Jones committed plagiarism."^{[Source:Smith]} In order to point out what the actual synthesis was, the A and B statements have to be presented in a completely unbiased way -- then twisted together in a way that advances a position.--Father Goose (talk) 06:03, 3 March 2009 (UTC)

I'm not sure if we're saying the same thing or not. Choosing a particular piece of article content to include, or a particular source for it, is not original research. That's garden variety sourcing. It may be problematic in other ways, i.e. that it is arbitrary or irrelevant to the topic at hand. If the second sentence, though sourced, is not sourced as applying to the first sentence, it is simply a non-sequitur but not original research in its own right. The actual OR is the synthesis of the two sourced statements to reach a conclusion not made in any source. I think I see the point that in our example as written there are two sins committed - the first by adding an impertinent definition of plagiarism and the second by using it to make a judgment about the Jones quote. We could reduce it to a single act of synthesis by adding a permissible second paragraph sourced as applying to the situation at hand: According to a New York Times article on the controversy, plagiarism in this case is defined as "xyz".[28] That sentence, standing alone, is fine. But it would still be synthesis to add a further sentence concluding that Jones committed plagiarism by the NYT definition, unless the NYT reached that conclusion. I hope that's clear. Wikidemon (talk) 08:12, 3 March 2009 (UTC)

Yep. Looks like everyone understands each other now.--Father Goose (talk) 17:45, 3 March 2009 (UTC)

It would be nice to have an example in the policy where some published material would be relevant to the topic, but not directly related to it, and where no obvious conclusion would be stated (non-syn). For instance, here is a nice real-life example from WP:NORN. --Phenylalanine (talk) 04:02, 5 March 2009 (UTC)

I can't see how this is an instance of SYN myself. SlimVirgin ^{talk|contribs} 07:19, 6 March 2009 (UTC)

That's the point. It's not synthesis, but it's still original research. I've seen countless times editors arguing that if material is "relevant" to the article topic, it cannot be original research. They essentially believe that "directly related" means "relevant". I don't blame them, since we haven't provided any explanations or examples in the policy to clarify what "directly related" means in practical terms. --Phenylalanine (talk) 11:22, 6 March 2009 (UTC)

We simply mean related to the topic according to the source; that is, the source must be writing about the topic of the article. But that's more to do with SYN than other forms of OR. What kind of OR do you see your example as if not SYN?

Personally, I'm not sure I see it as OR, though I can see what you mean. Asking that the source be writing specifically about the modern idea of the "Paleolithic Diet" is splitting hairs somewhat, in my view. The source was writing about what was eaten during the Paleolithic period.SlimVirgin ^{talk|contribs} 14:22, 6 March 2009 (UTC)

My example is equivalent to simply saying (to use the Smith & Jones example): "However, plagiarism is defined as using a source's information, ideas, words, or structure without citing them," which would be OR (but not syn) on your own admission above. And, in the same way that, in the Smith & Jones example, "The topic is not plagiarism. The topic is that particular dispute.", in the Paleolithic diet example, the topic is not the ancestral paleolithic diet, but the argument made by advocates of the modern diet that it is healthy because it resembles the ancestral Paleolithic diet. --Phenylalanine (talk) 01:13, 7 March 2009 (UTC)

I propose an addition to the primary sources policy:Proposal rescinded. Bongomatic 06:18, 14 March 2009 (UTC)

Our policy: Primary sources that have been reliably published (for example, by a university press or mainstream newspaper) may be used in Wikipedia, but only with care, because it is easy to misuse them. Any interpretation of primary source material requires a reliable secondary source for that interpretation. Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is verifiable by a reasonable, educated person without specialist knowledge. For example, an article about a novel may cite passages from the novel to describe the plot, but any interpretation of those passages needs a secondary source.

In certain contexts, primary sources are preferred for verification of certain isolated facts, where the source is primary only with respect to the fact, and not the topic of the Wikipedia entry. For example, membership in a hall of fame, receipt of an award, membership in a society, etc. are best demonstrated by reference to the granting institution's list of recipients. Secondary coverage of such isolated facts (from reliable sources) are nonetheless acceptable.

What do people think about this? Bongomatic 02:02, 9 March 2009 (UTC)

I'm not an expert in PSTS issues, but I think that in such cases, both primary and secondary sources are acceptable. Can you explain why you think primary sources are preferable in such instances? Cheers, Phenylalanine (talk) 02:49, 9 March 2009 (UTC)

Because you short-circuit potential errors or amplification of conventional wisdom where the (otherwise reliable) secondary source may have dropped the ball in fact-checking. Bongomatic 03:47, 9 March 2009 (UTC)

This is primarily a WP:Verifiability#Reliable_sources and WP:RS analysis, rather than a WP:PSTS issue. If the primary sources are useful and are assessed to be reliable, one should never hesitate to use them in keeping with the limitations described in WP:PSTS. ... Kenosis (talk) 03:10, 9 March 2009 (UTC)

If it weren't for the language here, it would indeed be primarily an issue for consideration on a case-by-case basis with respect to various reliable sourcing issues. However, on its face, this policy appears to control and lead people the wrong way. Bongomatic 03:47, 9 March 2009 (UTC)

The notion that this policy leads people "the wrong way" seems to imply a judgment that primary sources are inherently preferable. The language of WP:PSTS is designed to favor secondary sources because this is an encyclopedia and not a publisher of primary sources, nor of original research based upon primary sources. If the judgment is that a primary source is most reliable for the purpose you're involved in, such as reciting who won what award, use the primary source without apology, in keeping with what the policy says. If you're going to interpret the stuff in the primary, use a secondary source for such interpretation, criticism, analysis, etc. Seems pretty straightforward to me. ... Kenosis (talk) 04:44, 9 March 2009 (UTC)

My judgment is that primary sources are inherently preferable for certain types of information&mdash'a judgment that I believe (a) should be widespread; and (b) should be reflected here. Here is an example of why. Numerous reliable sources (the Associated Press, The Detroit Free Press, the Canadian Broadcasting Corporation, The Vancouver Sun, CanWest News Service, the Canadian Press) all reported Colleen Howe as an "enshrinee" into the United States Hockey Hall of Fame. But inspection of their website demonstrates that she is not.

While you are correct that nothing prevents an editor from determining that primary sources are preferable for a certain citation, nothing encourages that determination in the current policies or guidelines, and that seems to me to be (as I said) sending them "the wrong way". Seems pretty straightforward to me. ... Bongomatic 01:32, 10 March 2009 (UTC)

At the least, the provision must be that the primary source is published and easily accessible. I think we might make a rule that an easily accessible primary source should be also given when possible. But the example you give is a tricky one. In general, it can be non-trivial to say that a person is not on a long list, and more so if it involves conducting a search. It is quite common for people to miss it, because the name or the year is wrong--i have sometimes been able to verify in such cases only after very detailed research-level examination, especially if its a negative result. In this particular case, there must be some explanation--the news sources are presumably using some other designation than the one on the page you cite. The article on her you link to gives what you apparently think to be the story--but what you say there is really a matter of OR. Myself,I think its a matter of how the language is used: "enshrined" is the term in the official website, but "inducted" can apparently sometimes meet something much broader, or at least the sports news people seem to have thought so. This indicates the pitfalls. In a case like this, i'd look for an secondary discussion of the issue. DGG (talk) 01:51, 10 March 2009 (UTC)

I agree in principle with Bongomatic, but it has been hard to word the thing in a way that won't encourage the poor use of primary sources, as well as the good use of them. Bear in mind that a huge amount of primary-source use on WP is inappropriate. I hesitate to say "most," but it really is an awful lot.

But often a primary source is infinitely better. For example, I recently saw on Peter Singer that a quote was being attributed to him where he said that killing an "infant" is always less serious than killing a person, which I knew he would never have said. I checked the source, and it was the New York Times, quoting him from a talk he gave somewhere. I therefore went to his page at Princeton to see if I could find what he really said, and it was that killing a newborn baby is not as serious as killing a person who is able to know that they want to stay alive (whereas a newborn has no concept of that). So I changed the quote and added Singer as the primary source. Either the New York Times got it wrong, or Singer mis-spoke when he gave his talk (my guess is the former).

We could add something about it being sometimes appropriate to check with the original source to ensure secondary sources have reported it accurately. SlimVirgin ^{talk|contribs} 02:09, 10 March 2009 (UTC)

Interestingly, Jimbo has just spoken on the subject of using primary sources. His take:

In general, a book published by a person stating something about their [sic] personal life should be considered a reliable source - but there can of course be exceptions and complications. Judgment is always necessary.

In other words, unless there is good reason not to, primary sources should be considered reliable when people are offering details about themselves. --LK (talk) 09:24, 10 March 2009 (UTC)

It's highly qualified, and doesn't really affect policy. If they make credible or unobjectionable claims, they should be believed. Other claims require reliable secondary sources. Jayjg ^(talk) 20:53, 10 March 2009 (UTC)

It is indeed highly qualified. Immediately comes to mind James Frey. Best practice is to apply the overall core policies, NPOV, V and NOR. Where WP:V#Reliable_sources conflicts with WP:PSTS, simply make a sensible editorial judgment and use WP:PSTS as it's written, sticking to the facts or assertions that the primary source puts forward and citing to that primary source, using also the secondary sources to support the analysis of that material. In this case what we have is an issue brought up by Bongomatic w.r.t. Colleen Howe. If I'm reading it right (see the recent edit history of the Colleen Howe article) Bongomatic appears to possibly have caught a subtle error widely reported by the news media wherein Ms. Howe was given an award by the United States Hockey Hall of Fame (not to be confused with the Hockey Hall of Fame) and not actually inducted into that Hall. Well, to be frank, it's very good research that borders on original research but certainly doesn't cross the line. Based solely on this incident and any similar ones, I wouldn't want to see WP:PSTS rewritten to encourage WP users to prefer primary sources over secondary and tertiary sources. The rules already liberally allow the kind of caveat presently offered in the article on Colleen Howe w.r.t. what actual award she was given by the United States Hockey Hall of Fame, so there's no demonstrated need to explicitly encourage such research in WP:PSTS. Far more important for WP in general, IMO, is to limit fringe and quack theories by insisting that primary sources only be used to make factual assertions that don't take an astrophysicist or quantum mechanic or molecular biologist to double check if the statements in the article properly reflect what's said in the primary source. ... Kenosis (talk) 05:04, 11 March 2009 (UTC)

You guys are 'interpreting' the sense away from Jimbo's words. It's right there. A primary source is is a reliable source for information about themselves, unless there is good reason to think otherwise. LK (talk) 06:13, 11 March 2009 (UTC)

One of the reasons secondary sources are often preferred, even above primary sources writing about themselves, is the notability aspect. Is the information worth adding to the article? People tend to think everything about themselves is fascinating, but for us, the question is: does any secondary source agree? Another question related to secondary over primary is: is it fair to add the information to the article? Classic OR would be, for example, tracking down court documents (primary source material) to add details about a messy divorce to someone's bio. Again, we need a secondary source to tell us whether the divorce is something other people have decided is worth writing about — a Wikipedian deciding that on his own is OR. SlimVirgin ^{talk|contribs} 06:45, 11 March 2009 (UTC)

RE LK's (LawrenceKhoo's) quote of JWales's talk-page statement "A primary source is is a reliable source for information about themselves, unless there is good reason to think otherwise" : This, IMO, is always fair enough, if it passes muster with WP:V#Reliable_sources and ideally also passes muster with the related guideline WP:RS. The implication that WP:PSTS might need to be rewritten to support positions such as Bongomatic's issue in the article on Colleen Howe is, IMO, quite unnecessary. I leave aside here many of the additional implications having to do with user statements that seem to imply that User:Jimbo's talk page comments might somehow automatically be taken as policy, without any additional reference to everything else Wales has said throughout WP's history to date. Wales's authority w.r.t. editorial policy mainly derives from assertions made via the WP Foundation and to a lesser extent in other official advisory capacities, not from his many various personal user talk-page comments within the WP consensus process. In any case, Wales made it clear in his relevant talk-page comment that he didn't have enough information to make a judgment of any kind except to generalize his opinion that stuff people publish about themselves are reliable sources, absent evidence to the contrary. I think there's good reason to dispute that position, based on many more instances than, e.g., who's been actually, officially, legally and without a shadow of a doubt truly inducted into the US Hockey Hall of Fame rather than merely being a recipient of the Gretsky Award by the US Hockey Hall of Fame. ... Kenosis (talk) 07:11, 11 March 2009 (UTC)

The original suggestion had nothing to do with self-published information, and this conversation doesn't really seem to be tracking the original proposal (Jimmy Wales's unrelated comments on primary sources notwithstanding). This is about suggesting a preference for primary sources independent of the subject of an article for verifying isolated factual claims. As commentators have noted, such citation is clearly not prevented by current policy. However, it is equally not encouraged, and I think it ought to be. Bongomatic 01:06, 12 March 2009 (UTC)

The problem with finding the best words on this matter -- or on any description of Wikipedia policy -- is that individuals will wikilawyer the wording in order to promote their own agendas. The general rule in scholarship, here, is to always favor a primary source over secondary or tertiary ones. This is a rule I believe which works on Wikipedia more than 80% of the time. However, there are primary sources whose meaning or intent is disputed, where secondary sources must be used: for example, any article on the US Constitution. I believe that I'm not the one who doesn't want to read what some random person thinks the Second Amendment means, but what the experts & notable commentators think. And the best source for knowing what (to give a hypothetical example) the head of the NRA thinks it means is what he writes, not a secondary source that states what he thinks. We get ourselves tangled up in complex logical arguments because individuals demand (for good & bad reasons) we construe policy in a mechanistic manner, instead of with the spirit of its intent. -- llywrch (talk) 16:36, 12 March 2009 (UTC)

I think that's incorrect. This becomes especially problematic in WP:BLP's. I've seen biographies here where half the article is the subject's drunken 3AM blog post. And, yeah, sure they said that, but an editor cherry picking that quote to represent the subject's entire world view is very problematic. We really need secondary sources to attest to the notability of our information, otherwise it becomes indiscriminate. -- Kendrick7^talk 16:47, 12 March 2009 (UTC)

I think people often miss the forest for the trees when it comes to PSTS... the intent is to say "Don't do original research, and be extra careful about this when using primary sources". The intent is not to say "don't use primary sources where appropriate".

Of course there are times when citing directly to a primary source is both appropriate and preferred - such as a statement as to what the Second Amendment of the US Constitution says (ie quoting it) and there are times when citing to a secondary source is preferred (and where citing to the primary source would be inappropriate) - such as a statement as to what the Second Amendment means (ie how it should be interpreted). The key is that, while we can cite primary sources when appropriate, we have to be extra careful not to misuse them. In most cases where there is a choice, a secondary source is preferred. Blueboar (talk) 19:08, 12 March 2009 (UTC)

Kendrick, we already have a guideline about that --WP:UNDUE. Any article that is heavily based on anyone's drunken 3AM post is in trouble -- whether it is about a living, dead, or fictional person place or thing.

Blueboar, the problem with your preference is that too many people take it the next step & believe that if there are only primary sources available -- then nothing should be written. Even when they are used responsibly as, for example, an article I worked on today, Dale (woreda). I used lots of what some might consider primary sources in that article,; as a result they believe it is their duty to chop it into a stub, then nominate it for deletion because of a lack of reliable sources. (And if I sound paranoid about this possibility, it's not only because I deal with subjects on which there are too few secondary sources but because I've seen this exact thing happen.) -- llywrch (talk) 04:43, 13 March 2009 (UTC)

Broadly speaking, we shouldn't really have articles that are based entirely on primary sources. SlimVirgin ^{talk|contribs} 04:47, 13 March 2009 (UTC)

If you'd like to help me avoid this in regard to Ethiopian topics, I'd appreciate the help. Otherwise, I'm in the unusual -- & not always comfortable -- position where it is easier for me to get access to primary sources on my subject than secondary ones. With a few clicks on my keyboard, I can read the memoirs of a 19th century African explorer; if I want to read an expert discussion of him or his writings, I need to go thru Interlibrary Loan & wait a couple weeks for a copy of that expert's work to arrive, then spend money photocopying it if it proves to be useful. -- llywrch (talk) 16:20, 13 March 2009 (UTC)

WP:UNDUE doesn't really cover it, and if you think it does so you are missing my point. Even if an article is otherwise robust, a wikipedian would still be conducting original research by cherry picking selective quotes from the canonical works of the subject via their own personal volition. There needs to be a reliable secondary filter for such information or else, as best I can tell, all hell breaks loose. -- Kendrick7^talk 10:28, 13 March 2009 (UTC)

If it doesn't cover it, then it should. I agree with you that cherry-picking material violates the idea of an honest article. However, that can be done with primary and secondary sources; a skilled writer knows many ways to introduce an unfair bias into an article, & "cherry-picking" is just one of them. My point was that instead of wrangling the language to prevent all possible ways to introduce an unfair bias, we interpret the language with the intent that this is our goal. -- llywrch (talk) 16:20, 13 March 2009 (UTC)

I will again try to turn this discussion back to the original proposal in this thread. This is not intended to be a proposal about primary sources in general, and specifically excludes primary sources that are the subjects (or are closely linked) to the subjects of the articles in which they are cited. Rather, it is intended to encourage something that is unambiguously permitted by policy but unintentionally discouraged. This is not about the ramblings of the subject of a BLP, or about some editor's interpretation of the Second Amendment, referenced with a citation to the text alone. This is about specific, verifiable facts used in relation mainly to people (awards, etc.), songs and albums (chart positions), books (position on best-seller lists), etc., where—because the subject has already been demonstrated to be notable, so secondary evidence of that has already been established with other references—there is no benefit of the synthesis provided by the secondary sources (again, for such isolated facts), and the primary source significantly reduces the possibility of error or received bias in reporting.

• SlimVirgin: No suggestion at all that all primary sources could be appropriate.
• Kendrick: No cherry picking, as not for supporting opinions, lines of reasoning, or conclusions at all—just isolated facts.
• Blueboar: Just so—suggesting that this interpretation (although much more limited in scope than your formulation) should be encouraged.

Bongomatic 11:45, 13 March 2009 (UTC)

Perhaps there should be due mention of suitable places where Original Research (of sufficient quality, naturally) can be placed, general or specific to the subject (we all know several).

A version of Marx' comment on Feuerbach - that Wikipedia describes the world, and other sites should be used to research and change it might be appropriate. Jackiespeel (talk) 15:05, 13 March 2009 (UTC)

I didn't think that there are any exceptions to No Original Research. A Quest For Knowledge (talk) 15:21, 13 March 2009 (UTC)

TBH, it's I see no evidence that primary sources relating to "bestseller lists" and "awards lists" are inherently more reliable or preferable than secondary sources for isolated facts of the kind illustrated at the article on Colleen Howe, at least not in a way that can be generalized as an explicit category of primary source. It's a WP:V#Reliable_sources and WP:RS analysis that should be implemented on a case-by-case basis. If evidence is found that a source is in error, it's totally appropriate to attempt to use one or more other sources to correct it. For instance, the weight given to the list of "enshrinees" into the United States Hockey Hall of Fame w.r.t. Colleen Howe is given to it not because it's a primary source, but because we're inclined to deem the USHHF the most reliable source of who their enshrinees are. If the USHHF were demonstrated to be often careless or erroneous with their online lists of enshrinees, we'd look to other RSs to make that judgment. Again, it's an issue of reliability, not of whether the alleged "fact" is published by a primary source. Quite commonly what comes "from the horse's mouth" is "primary" but nonetheless unreliable. ... Kenosis (talk) 16:34, 13 March 2009 (UTC)

Kenosis hits the nail right on the head. Wikiepdia does not ban the use of primary sources. Yes, we limit their use... because it is easy to misuse primary sources in ways that constitute Original Research. However, there is nothing wrong with using a primary source, if it is the most reliable source for the information you are presenting and what you are presenting is not OR. Remember why we caution against primary sources, and then use your judgement as to whether to use one.

There is an unfortunate tendency, when editors are in a dispute over something, to want to add a clarification to policy to cover the specific issue at hand in the dispute. This is a natural desire (I would be willing to bet that most of us first got involved in editing on policy pages and helping on the various noticeboards for exaclty that reason... I know I certainly did), but it is a desire that should be resisted. It leads to instruction creap, and actually makes our policies harder to understand. It leads to a legalistic narrow interpretation of our polices and guidelines (and then to wikilawyering over the meaning of individual policy statements instead of understanding of the broader intent of the policy). If you have not read it before, go readWikipedia:The rules are principles, a wonderful essay on how to approach Policy. Focus on the intent of of our policies and guidelines. Then apply that intent to each individual article. Blueboar (talk) 17:22, 13 March 2009 (UTC)

Well said, Blueboar. SlimVirgin ^{talk|contribs} 00:16, 14 March 2009 (UTC)

Ditto. Nonetheless I want to say that IMO it was very good research on Bongomatic's part in seeking best possible accuracy w.r.t. the topic of Colleen Howe. I imagine it likely that this quest for accuracy applies to other topics in which s/he is involved as well, and is quite helpful to "the project"... Kenosis (talk) 04:52, 14 March 2009 (UTC)

The proposed change wasn't supposed to be a big deal—just a clarification. In fact, I have always sought primary sources for awards and the like (the Emmys are particularly frustrating in this regard), and have never (as suggested above) had anyone criticize the addition of such references as contradicting policy. However, the Colleen Howe discovery made me realize how useful it would be if people (who bother to read policies!) generally sought out primary sources in this sort of situation. I haven't considered the ways in which the proposed language could be hijacked by POV-pushers, but tried to tailor it narrowly. To me, the two features of the proposal that make it relatively unlikely to be used for unintended purposes are that (1) the "preferred primary sources" would never be primary with respect to the subject of the article; and (2) such sources would be cited for isolated facts, not interpretations or opinions of such facts. But, I don't really care whether it's adopted. I'm surprised it generated so much discussion (and even more surprised about how vanishingly little of that discussion was on topic). Bongomatic 06:17, 14 March 2009 (UTC)

Flights of fancy

Labyrinthine policy wording cannot take the place of common sense. Primary sources have their place as long as we do not attempt to interpret them. Simple as that.

A fine example came along other day as I worked on a new restoration of the Wright brothers' first flight. The same group of images at the Library of Congress had several scans of primary documents, including one with the curious title "M.F. Keiter "blackmail" letter".[13] Blackmail? What could that be? Well, being a normal human being with a weak spot for gossip (even century old gossip about the squeaky clean Wright brothers) I downloaded the thing and had a look. Indeed it was a blackmail letter. And there were several others like it in the collection. A Google Books search turns up mentions of Keiter. Turns out from secondary sources that an M.F. Keiter was the villain in a financial scandal and the Wright brothers' father (a bishop in a Protestant church) went out of his way for several years in an effort to bring the matter to justice. For a while it looked like this might be a DYK in the making, but the limited resources at Google Books don't mention this blackmail. They do mention a different blackmail a couple of years later when another unscrupulous fellow tried to coerce a share of the profits from their invention of the airplane (pity the poor upstanding Wright family), but that was entirely unrelated to a fraud in the church printing press bookkeeping. So if anyone feels like looking up more sources that discuss how nasty this particular scandal became, go ahead. We've got scans of original blackmail and forgeries to supplement those secondary sources, if secondary sources turn up. Otherwise--if you delve the nasty little church scandal--don't publish here first. Find a reliable venue and write up your exposé in a reliable source, along with any analysis you can get past the editor, and then Wikipedia can cover the story. Other Wikipedians could still use those original sources as supplements; we just can't connect the dots. Durova^Charge! 05:18, 15 March 2009 (UTC)

Yes indeed. I think the current language of WP:PSTS, though debatable and often debated, remains consistent with the assertion that "Labyrinthine policy wording cannot take the place of common sense", and that "Primary sources have their place as long as we do not attempt to interpret them." . At least I hope so. ... Kenosis (talk) 06:17, 15 March 2009 (UTC)

I've removed these from the routine calculations section, as both can be problematic in terms of OR and POV. We should stick to the rule that reliable sources should have included these before we use them — emphasizing a percentage over a number can paint quite a different picture, as can introducing a graph.

In general, I think we need to be careful not to state the obvious in this policy. It has traditionally been quite tightly written so that it's easy for editors to scan quickly. SlimVirgin ^{talk|contribs} 23:59, 2 March 2009 (UTC)

Again you removed parts, or similar parts, regarding graphs and percentages that were restored twice before by two different editors. You need to get consensus on this before removing them again. Your remarks above are unclear and you didn't even wait to get a response before you removed those items again, over the restorations by two editors.

Regarding the rest of your edit, which had to do with the Translations section, there is currently a discussion going on regarding that section here. It appears that your edit has the same problems that were brought up regarding that Translations section and there is a proposed improvement given in that section with explanation. The basic purpose of the improvement is to use WP:NONENG as the Wikipedia source of the info, rather than to explain it here, where it is incomplete and possibly inconsistent with WP:NONENG, either now or in the future after edits are made on both policy pages that may change their respective info re non-english sources.

For the above reasons, I undid your edit. Please clarify your above remarks by explaining them in more detail, make your case at the discussion section on Translations, and get a consensus before attempting to reintroduce any of your edit into the article again. Thank you. --Bob K31416 (talk) 01:20, 3 March 2009 (UTC)

And now two different editors have removed it. Putting things into a graph is definitely not a "routine calculation". And I've seen areas where people have attempted to replace percentages with numbers precisely for the purpose of promoting a specific POV. Please get a wider consensus for this modification of policy. Jayjg ^(talk) 03:57, 3 March 2009 (UTC)

In general, user created images, including graphs, are not subject not the OR policy. — Carl (CBM · talk) 04:11, 3 March 2009 (UTC)

"In general"? Can you show where that is outlined in policy? Aside from the recent attempt to insert it here, I mean. Jayjg ^(talk) 04:16, 3 March 2009 (UTC)

"Original images created by a Wikipedia editor are not, as a class, considered original research – as long as they do not illustrate or introduce unpublished ideas or arguments, the core reason behind the NOR policy. ". This is the general practice on many articles. Editors are free to create illustrations of any sort, subject to general editing consensus. There is no need to show that a diagram, graph, or chart has been previously published before it can be used in an article. We encourage editors to create new illustrations and release them under a free content license. — Carl (CBM · talk) 04:24, 3 March 2009 (UTC)

Why do you think graphs are included under "images"? Jayjg ^(talk) 04:28, 3 March 2009 (UTC)

Because they are in the Image namespace. — Carl (CBM · talk) 04:32, 3 March 2009 (UTC)

That's a technicality, as Wikipedia has no graphing capabilities. I could put anything I wanted into an image, but that wouldn't mean all items in the Image namespace were exempt from the WP:NOR policy. Jayjg ^(talk) 04:39, 3 March 2009 (UTC)

Yes, the entire image namespace is in a certain way exempt from NOR, which is mainly focused on the textual content. Original graphs, diagrams, charts, photographs, etc. are encouraged, as we are not permitted to use non-free copyrighted images in most cases. Of course the data used to create the graph needs to be verifiable, but the graph itself need not be previously published before we can use it. Indeed, if it were previously published and non-free we would be unable to use it or any derivative work of it. — Carl (CBM · talk) 04:42, 3 March 2009 (UTC)

Where does it say that "the entire image namespace is in a certain way exempt from NOR"? If I include a new theory in an image, is it suddenly no longer NOR? Jayjg ^(talk) 04:50, 3 March 2009 (UTC)

"Original images created by a Wikipedia editor are not, as a class, considered original research – as long as they do not illustrate or introduce unpublished ideas or arguments, the core reason behind the NOR poliy." Images means... images. WP:V and WP:NPOV apply to all content, but WP:NOR permits original images. — Carl (CBM · talk) 04:54, 3 March 2009 (UTC)

The source you quote says nothing about "the entire image namespace", and your argument is circular. Whether or not graphs are considered part of the "class", and whether or not they "illustrate or introduce unpublished ideas or arguments", is the issue being debated here. Jayjg ^(talk) 04:57, 3 March 2009 (UTC)

Images are, by definition, things in the image namespace. Provided that a graph is based on verifiable data, the mere fact that the graph itself was not previously published (which seems to be the issue at hand) is specifically not a concern of the NOR policy. — Carl (CBM · talk) 05:05, 3 March 2009 (UTC)

Also, edit warring over the content of the policy is hardly best practice. [14]. — Carl (CBM · talk) 04:26, 3 March 2009 (UTC)

Huh? Please review WP:BRD. Jayjg ^(talk) 04:28, 3 March 2009 (UTC)

BRD is often mistaken as a license to edit war. It is not. The material was reverted once; reverting it repeatedly, in this sort of context, is edit warring.

Also, the second revert here is inappropriate. [15] [16]. Let's all try to follow best practices when we tell other people about them. — Carl (CBM · talk) 04:32, 3 March 2009 (UTC)

You seem to have the entire process backwards. Two editors tried to re-write longstanding and fundamental Wikipedia policies. One of them was an editor of less than a year, and under 1000 article edits; the other had under 1800 article edits. The change was reverted, and that should have been the end of it, until a much wider consensus was reached. Please reserve your admonitions for those to whom they are more appropriately directed. Jayjg ^(talk) 04:39, 3 March 2009 (UTC)

Admonitions are most appropriately directed at editors with the most experience, and administrators, as they are expected to reflect the core principles of the site. Newer editors may be less familiar with these principles. — Carl (CBM · talk) 04:44, 3 March 2009 (UTC)

CBM, you can't have it both ways. Either they're new editors, in which case they shouldn't be modifying fundamental policies, or they're experienced editors, in which case they should abide by WP:BRD. Which is it? Jayjg ^(talk) 04:50, 3 March 2009 (UTC)

Jayjg, in this case I was referring to experienced editors who should abide by WP:EW. — Carl (CBM · talk) 04:54, 3 March 2009 (UTC)

Got it. Well, Carl, one single revert of changes to fundamental policy by inexperienced editors does not constitute an "Edit War". Jayjg ^(talk) 04:57, 3 March 2009 (UTC)

I stand by my original comment. — Carl (CBM · talk) 05:05, 3 March 2009 (UTC)

This is an interesting issue, IMO, though it might not need to be all that complicated from a policy standpoint. Take, for just one of countless instances, United States public debt which features File:USDebt.png. Click on the file page to see the prior forms of this file on the file page. Even despite its careful sourcing to a "reliable source" (the White House Office of Management and Budget), it has taken on different forms through its history. And yet, in each case it appears to faithfully graph the data from which it is derived. It appears to me that the most relevant issues with a graph are NPOV and V., and that if a local article consensus is that a graph meets NPOV and V, then it's likely not original research or original synthesis. SlimVirgin's point about not having been explicitly needed before is fair enough, except that we've recently had some vociferous complaints about the implication that NOR forbids routine calculations. To the extent that a graph is a set of routine representations of existing numerical data points, not even calculations per se, it seems fair enough to me to explicitly allow them so long as they're verified to one or more RS's and are consensused not to be synthesis advancing a POV. Or have I missed some other important aspect of this issue? ... Kenosis (talk) 04:46, 3 March 2009 (UTC)

If people want to add to this policy that the creation of a graph isn't OR, it will need widespread consensus. It really shouldn't be added (and reverted over objections) just because a couple of editors on the page feel it's okay. Better to say nothing at all about graphs, and treat each case on its individual merits, than to say explicitly that graphs constitute exceptions to the OR policy. SlimVirgin ^{talk|contribs} 05:23, 3 March 2009 (UTC)

As I pointed out, the policy already says that images are not subject to the OR policy provided they are based on verifiable data. There is no need to change the policy to achieve this. — Carl (CBM · talk) 05:30, 3 March 2009 (UTC)

As I pointed out, your argument is circular. Putting something in the image space doesn't immediately immunize it against being Original Research. I could create an image of text, and upload that too, but that wouldn't mean the text wasn't original research. And I have, indeed, in the past, argued that images (and I mean real images, not complicated calculations like graphs) were exempt from the WP:NOR strictures, but was quite forcefully told the opposite. Jayjg ^(talk) 05:33, 3 March 2009 (UTC)

Jayjg, Please note that the item at issue was put in last Christmas here by an editor other than the two that tried to restore it over SlimVirgin's recent two deletions of it. --Bob K31416 (talk) 05:32, 3 March 2009 (UTC)

Wow, you mean the change is all of 2 months old? Did that editor get any consensus for the change before doing so? Jayjg ^(talk) 05:36, 3 March 2009 (UTC)

Where have you been in the last two months? There should be some stability to the editing process. You and SlimVirgin could just as well have had a discussion before deleting it. Who knows, maybe people would agree with you if you explained it calmly. This is not calm, nor orderly, nor considerate of the other editors who have been working here during the last two months since the item was added. --Bob K31416 (talk) 05:40, 3 March 2009 (UTC)

The policy has existed for what, over 6 years now? And it's had the addition of "graphs" for 2 months? I suggest that the insertion of that term, based on the input of 3 editors, two of them fairly inexperienced, was inappropriate for one of Wikipedia's fundamental content policies. Jayjg ^(talk) 05:45, 3 March 2009 (UTC)

As I understand it, the question here is about verifiable data (for example, a table of numerical data) that is converted into graphical form. My point is that this sort of conversion is not subject to the OR policy simply because the form of the graph is new. We encourage editors to create new illustrations, figures, and graphs for articles, without the requirement that these have appeared in sources before. Indeed, because of the scarcity of pre-existing free images, we have to rely on Wikipedia editors making new free content if we hope to illustrate our articles.

There may be NPOV concerns with a graph, and these should be taken seriously. However, if the graph is based on verifiable numerical data, there is no concern with the NOR policy. — Carl (CBM · talk) 17:03, 3 March 2009 (UTC)

Nicely put. Axlrosen (talk) 02:44, 5 March 2009 (UTC)

Are there any remaining objections to putting the graph part back into Wp:NOR? And if so, please specify what they are. Thanks. --Bob K31416 (talk) 18:53, 11 March 2009 (UTC)

Yes, I object. Graphs can be problematic for various reasons, including introducing (or appearing to introduce or over-emphasize) a POV. It's better not to mention them and let editors judge on a case-by-case basis. If we encourage them here, graphs will appear whether they're appropriate or not. SlimVirgin ^{talk|contribs} 00:10, 12 March 2009 (UTC)

Regarding your remark, "Graphs can be problematic for various reasons, including introducing (or appearing to introduce or over-emphasize) a POV." -

CBM answered the concern regarding POV, in a discussion in which you participated. You didn't respond to CBM's remark that was made 8 days ago and which hasn't been challenged. CBM essentially noted that POV isn't relevant to WP:NOR, when he wrote, "There may be NPOV concerns with a graph, and these should be taken seriously. However, if the graph is based on verifiable numerical data, there is no concern with the NOR policy."

Please specify any problems relevant to WP:NOR that you are aware of regarding graphs. Thank you. --Bob K31416 (talk) 00:55, 12 March 2009 (UTC)

Since SlimVirgin didn't specify any problems with graphs that are relevant to WP:NOR, and since the phrase regarding graphs is only a clarification of existing WP:NOR policy in WP:OI, along the lines that CBM mentioned regarding images, and graphs are a summarizing of info from reliable sources, which is what good Wikipedia editors do, I have added the phrase "nor forbid displaying data in a graph". --Bob K31416 (talk) 03:10, 15 March 2009 (UTC)

Actually, SV quite clearly specified problems with graphs that are relevant to WP:NOR. In addition, I have pointed out that they are not merely simple calculations or summarization. The fact that CBM responded to SV's point does not mean that he addressed it - those are two different concepts. In addition, the argument that graphs were merely images and therefore exempt was circular. And finally, a significant loophole in policy is not just "a clarification of existing WP:NOR policy in WP:OI", and it is disingenuous to claim it is. I am again removing this change to policy. Please get wider consensus for this policy change. Jayjg ^(talk) 01:41, 17 March 2009 (UTC)

This isn't actually a change to policy. People add graphs to articles quite often, and moreover images are not subject to the OR policy, as the policy already says and as I pointed out above. "Original images created by a Wikipedia editor are not, as a class, considered original research – as long as they do not illustrate or introduce unpublished ideas or arguments, the core reason behind the NOR policy."

So the clarification isn't really needed, but it doesn't change the actual practice which permits users creating original graphs. Finally, could you point out which comment by SlimVirgin actually addresses NOR, rather than NPOV? — Carl (CBM · talk) 01:47, 17 March 2009 (UTC)

You've made those statements, and I've pointed out that your argument regarding graphs being images is circular. Moreover, if the issue is as cut-and-dried as you claim, then there's no need for a special phrase to allow it. If the "the clarification isn't really needed", then let's not bloat policy by adding it. Jayjg ^(talk) 01:56, 17 March 2009 (UTC)

How is it circular to point out that the NOR policy directly says that it does not apply to original images? Note I have not edited the policy page about this, I have only expressed my opinion here. — Carl (CBM · talk) 02:05, 17 March 2009 (UTC)

Again, as pointed out, graphs aren't "images". And if you state they are saved as image files, I have already pointed out that anything can be saved as an image file, including text, but that does not make the text in the image file exempt from the NOR policy. Note I have never claimed you edited the policy page about this. Jayjg ^(talk) 02:12, 17 March 2009 (UTC)

Also, Jayjg and Bob K, I already pointed out above that it's inappropriate to keep edit warring over this. If there is really widespread objection to the change someone else will revert it; you each already had your go at it. — Carl (CBM · talk) 01:51, 17 March 2009 (UTC)

Hmm. As I've pointed out, your suggestions are better addressed to those trying to edit-war in their modifications to policy. Policy changes require significant consensus. If you still want to change this policy, why don't you try getting a real consensus for it? I'll certainly go along with whatever the consensus turns out to be. Jayjg ^(talk) 01:56, 17 March 2009 (UTC)

I have not edited the NOR page about this. You, however, continue to edit war over it, despite my pointing out that this is inappropriate. As you are a very experienced editor, my admonition about edit warring is primarily directed at you. — Carl (CBM · talk) 02:05, 17 March 2009 (UTC)

I haven't claimed you did edit the NOR page about this. I have not "edit-warred" over this, I have removed this non-consensus change to policy twice, the first time two weeks ago, and the second time today. Do not again admonish me for things I have not done; in fact, please restrict your comments to discussion of the NOR policy, which is what this Talk: page is for. I will ignore future comments that are not restricted to discussions of policy. Jayjg ^(talk) 02:09, 17 March 2009 (UTC)

Removing the same edit twice in this way is, by definition, edit warring. If you consult WP:EW you will see that "An edit war occurs when contributors, or groups of contributors, repeatedly revert each other's contributions." Moreover, as the text had already been reverted before your first revert, and was under discussion here, your first revert was also inappropriate (this is the "groups of contributors" language in WP:EW). Editors who are entrusted with positions of responsibility are expected to be particularly careful in modeling best practices, and you have not done so here. — Carl (CBM · talk) 02:14, 17 March 2009 (UTC)

Hmmmm. This is all very suspicious. --Bob K31416 (talk) 04:19, 17 March 2009 (UTC)

Err, what? Jayjg ^(talk) 04:21, 17 March 2009 (UTC)

Well that's darn interesting. I make a little remark and you reply immediately. But you don't reply to CBM's remark. Why is that? --Bob K31416 (talk) 04:30, 17 March 2009 (UTC)

Read my last comment to CBM, which explains exactly why. Now, what did you mean when you said "This is all very suspicious"? Jayjg ^(talk) 04:43, 17 March 2009 (UTC)

My oh my. You'd rather pursue this than respond to CBM's remark. Perhaps you should ask yourself some questions and do some soul searching. --Bob K31416 (talk) 05:27, 17 March 2009 (UTC)

Bob, I don't know what your issue is, but you seem determined to make a potentially major policy change over objections. Did you have problems with a graph somewhere yourself? As has been pointed out already, we're not saying unpublished graphs can't be used; we're just not encouraging them. That way, the editors on the page can decide in context. SlimVirgin ^{talk|contribs} 08:28, 17 March 2009 (UTC)

So now you respond, huh. Here's a trip down memory lane:

Why didn't you respond back then. I carefully prepared the ground for my addition and all I got was this current obstruction and no response from you. Me edit warring? What bunk. BTW I saw how you and Jayjg got the infamous example into the Synth section by edit warring.[17] That is apparently both of yours' style. You two are quite a team, much to the detriment of WP:NOR. --Bob K31416 (talk) 11:58, 17 March 2009 (UTC)

Category:Images of curves, and friends

Replying to an earlier post of Jayjg, because I think it's an important point, I cannot see how one can argue that graphs are not images, or are somehow subject to the OR policy in the sense that the images must already be published before we can use them. This runs completely against our actual practice, and the language of this page itself.

Re longstanding practice: we have entire categories such as Category:Mathematics images and commons:Category:Curves which consist of user created graphs. These categories seem to me to be a clear demonstration that user created graphs are permitted and welcome! We do not make users prove that their graph has already been published before they can use it; as I have pointed out, if the graph were taken from a nonfree source we could not use it anyway. If these original, user-created graph images were prohibited by the NOR poicy, we would not have categories full of them. — Carl (CBM · talk) 12:54, 17 March 2009 (UTC)

I agree that user created graphs are images, and as such fall under WP:OI... WP:OI states: "Original images created by a Wikipedia editor are not, as a class, considered original research – as long as they do not illustrate or introduce unpublished ideas or arguments, the core reason behind the NOR policy" and "Images that constitute original research in any way are not allowed." These restrictions apply to graphs no less than photographs.

In other words, a user created graph is fine - as long as it does not illustrate or introduce umpublished ideas or arguments... it must be based upon published data. While the graph may fall under WP:OI and thus not need to be sourced... the data underlying the graph is a different matter. All images should be a representation of something discussed in the text of the article... The same is true of a graph. And the text that it represents does need to be referenced to a reliable source.

Now, as to the edits to this policy that are at issue here... Since graphs are already discussed in the OI section, I see no need to mention them again in the Routine calculations section. Blueboar (talk) 14:12, 17 March 2009 (UTC)

I completely agree (I think everyone does) that the underlying data for a graph needs to be verifiable. — Carl (CBM · talk) 14:20, 17 March 2009 (UTC)

What's the point here? Jayjg and SlimVirgin will just edit war and keep it out. CBM wrote, "If there is really widespread objection to the change someone else will revert it". CBM had a chance to revert to support what he is currently advocating and he didn't. --Bob K31416 (talk) 14:47, 17 March 2009 (UTC)

I would revert it out as well. It isn't needed. Blueboar (talk) 16:02, 17 March 2009 (UTC)

You're right, WP:OI does mention diagrams and one of the definitions of a diagram is a graph.[18] I'm not sure how many people realize this, but a graph is a type of diagram, according to a dictionary.

SlimVirgin and Jayjg, Are you OK with that, i.e. that editors are encouraged to draw diagrams, including graphs? --Bob K31416 (talk) 23:22, 17 March 2009 (UTC)

As long as the diagram (in this case a graph) does not "illustrate or introduce unpublished ideas or arguments", it passes WP:OI. Blueboar (talk) 00:38, 18 March 2009 (UTC)

Do you think an editor who is coming to WP:NOR for help regarding graphs, will realize that they are OK from what is said in WP:OI? I don't think either of us realized before that one of the definitions of diagram was graph. --Bob K31416 (talk) 13:31, 18 March 2009 (UTC)

I think most people would... I did. Blueboar (talk) 13:41, 18 March 2009 (UTC)

Sure. Sorry for my incredulity but if the people posting messages in this section (including you) understood that one of the definitions of diagram was graph, it seems that they wouldn't have to go through a discussion re images. No one even mentioned the word diagram before I mentioned it. They could have just simply quoted the following sentence from WP:OI and note that a definition of diagram is graph: "Wikipedia editors are encouraged to take photographs or draw pictures or diagrams and upload them, releasing them under the GFDL or another free license, to illustrate articles." --Bob K31416 (talk) 13:52, 18 March 2009 (UTC)

Is math original research? Say a game sold 5.2 million in Japan, and 6.7 million in the rest of the world. Is it original research to say something like the game has sold 11.9 million copies worldwide. (A simple example, but still, the point is made) IAmTheCoinMan (talk) 11:41, 18 March 2009 (UTC)

No, the policy permits simple calculations such as that. Although there can be many other concerns if the two numbers come from different sources or something like that. But if a single source says, "The game sold 5 million in Japan and 6 million everywhere else", we can certainly attribute to that source the claim that the game sold 11 million. — Carl (CBM · talk) 12:12, 18 March 2009 (UTC)

And what if the numbers come from different sources. Say 1 source says Halo sold 13 million, another source says Halo 2 sold 7 million. Then overall, Halo and Halo 2 have sold 20 million copies.

IAmTheCoinMan (talk) 12:28, 18 March 2009 (UTC)

That is more problematical, and might be considered a WP:SYNT violation. The example you give is not that controvercial, and probably would not be challenged if you added it to an article, but if you are at all concerned that it might be challenged it is probably better to simply state that "Halo sold 13 million copies (cite to source 1) while Halo 2 sold 7 million copies (cite to source 2)" and not attempt to combine the figures. Blueboar (talk) 13:07, 18 March 2009 (UTC)

Original research is good! Fangoriously (talk) 16:58, 15 March 2009 (UTC)

Yes, it is... but not on wikipedia. Blueboar (talk) 21:49, 15 March 2009 (UTC)

Publish elsewhere in a reliable source. Then we can cite you. Unlike Wikipedia, you might even get paid for that. ;) Durova^Charge! 02:07, 17 March 2009 (UTC)

Why isn't there a wiki for original thought? --Fangoriously (talk) 16:09, 22 March 2009 (UTC)

There are... many. Wikia allows OR, as does Wikiversity. Blueboar (talk) 16:18, 22 March 2009 (UTC)

I think we need to adjust this policy (along with WP:CITE) so that VERY obvious stuff doesn't need citations.

Take for example: Warner Music Group](WMG) is the third-largest of the big four major record labels, the others being Sony Music Entertainment, EMI, and Universal.

This opening line is a very obvious statement, and thus already has no citation. And [19] is another obvious thing, and shouldn't need citations.

As of now, the policy as it is, is too strict. I think we can loosen it up a bit. FMAFan1990 (talk) 05:53, 22 March 2009 (UTC)

I think the policies and guidelines already address that (including WP:V). While citations are always encouraged, no citation is ever necessary if the thing being cited is not challenged and not likely to be challenged. COGDEN 06:11, 22 March 2009 (UTC)

In this specific case, how does one know that Warner Music Group is the third largest? How likely is it that the non specialist reader will know that it is third and not second or fourth? Also by what criteria is it third? Is it by stock market capitalisation? As a contrast FMAFan1990 do you also know of the top of your head which is the third largest oil company by stock market capitalisation? One might argue that the average reader of Wikipedia should know that Warner Music Group is a large music publishing house (so no need to cite a statement like "WMG is a large music publishing company"), although personally I do not think it is an something that we can expect a reader to know as an obvious fact like "the sun rises in the east" so I would expect it to be sourced somewhere in the article but not necessarily in the lead. -- PBS (talk) 11:02, 22 March 2009 (UTC)

Agree with both Cogden and PBS... this is really a WP:V issue and not one for NOR, and WP:V already clarifies when things need to be cited and when they don't. Blueboar (talk) 14:43, 22 March 2009 (UTC)

The following survey pertains to a section of WP:NOR that is titled Synthesis of published material that advances a position and is also known as WP:SYNTH.

---

The following example and discussion is currently in WP:SYNTH, today Mar 3, 2009:

The following example is based on an actual Wikipedia article about a dispute between two authors, here called Smith and Jones.

Smith claimed that Jones committed plagiarism by copying references from another author's book. Jones responded that it is acceptable scholarly practice to use other people's books to find new references.

Now comes the original synthesis:

If Jones did not consult the original sources, this would be contrary to the practice recommended in the Harvard Writing with Sources manual, which requires citation of the source actually consulted. The Harvard manual does not call violating this rule "plagiarism". Instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them.

The first paragraph was properly sourced. The second paragraph was original research because it expressed the editor's opinion that, given the Harvard manual's definition of plagiarism, Jones did not commit it. To make the second paragraph consistent with this policy, a reliable source would be needed that specifically comments on the Smith and Jones dispute and makes the same point about the Harvard manual and plagiarism. In other words, that precise analysis must have been published by a reliable source in relation to the topic before it can be published in Wikipedia by a contributor.

---

The following example and discussion is proposed to replace the above:

The following example is based on an actual Wikipedia article about a dispute between two authors, here called Smith and Jones.

Smith claimed that Jones committed plagiarism by copying references from another author's book. Jones responded that it is not plagiarism, but acceptable scholarly practice to copy references from other people's books.^[27]

Plagiarism is defined as using a source's information, ideas, words, or structure without citing them.^[28] Since Jones did not cite the book from which he obtained his references, but only listed the references themselves, he committed plagiarism.

Two parts of the above example are properly supported by sources, indicated by [27] and [28] respectively. The part highlighted in bold uses information from each of the previous two parts, to form a conclusion that is in neither of the sources. Furthermore, there is no other source indicated for the conclusion. It appears that the conclusion was conceived by the editor, and is thus original research prohibited by this policy.

---

Please indicate whether you support or oppose the proposed version and you may comment on your reason(s). Thank you. --Bob K31416 (talk) 09:08, 3 March 2009 (UTC)

---

• Support - the proposed change is clearer, in my opinion. --Bob K31416 (talk) 09:08, 3 March 2009 (UTC)
• This is a strange matter for a consensus vote, and an odd issue all the way around. The newer version, agreed to in a painstaking step by step, transparent process here on the talk page, is clearly better than the original (which is an intermediate version of the old example). It fixes a number of problems. However, there is an objection that the example posed of a synthesis contains more than one piece of OR. Thus, it could be improved. I agree - and welcome any suggestions regarding a more clear, straightforward, illustrative example of synthesis that could follow the first half of the example. If so, great. Let's do it. If not, objection to an improved version should hardly be taken as endorsement of the older one. Wikidemon (talk) 10:07, 3 March 2009 (UTC)

• Thank you for your comment. It would be helpful if you stated whether you Support or Oppose the proposed change. Thank you. --Bob K31416 (talk) 20:39, 3 March 2009 (UTC)

• Support... but... to be honest, I think we would do even better by replacing the Smith/Jones plagarism example with a completely different example. Something that is more obvious. —Preceding unsigned comment added by Blueboar (talk • contribs) 13:02, 3 March 2009 UTC
• Comment. This example, first added a two or three years ago by SlimVirgin, has served fairly well on the whole. But it has suffered from lack of maintenance and the "quote" using Smith and Jones in place of Finkelstein and Dershowitz has been altered substantially for extended periods of time. Then somebody comes along and says essentially "this makes no sense", and the original debate needs to be found again (as I've done twice in the last year) with the original quotes and the original sequence of events that gave rise to the example. So it needs to be better maintained and the links to the orignal scenario better preserved so they can be more conveniently accessed for reference. My only complaint about it is that it has a bit too much information, involving as it does the extra issue of plagiarism in the actual presented scenario, in a way that I think does not particularly help WP users to get an appropriate sense of the middle ground between original research on the one extreme and plagiarism/copyvio on the other extreme. IMO, it would be helpful if someone found an example that didn't have the extra potentially confusing issue of plagiarism such as was involved in the original scenario of Finkelstein and Dershowitz.
  ..... For the links to the actual original scenario involving the Dershowitz-Finkelstein affair, from which SlimVirgin derived this example, perhaps Bob31416 or Wikidemon can link to the appropriate spots in the archives--I've lost track of them at the moment. Thanks. ... Kenosis (talk) 16:17, 3 March 2009 (UTC)

• Thank you for your comment. It would be helpful if you stated whether you Support or Oppose the proposed change. Thank you. --Bob K31416 (talk) 20:39, 3 March 2009 (UTC)

• You can see the editing process (and links to the original discussion, etc) in the most recent archive Wikipedia talk:No original research/Archive 43#Problem with example of OR in WP:SYN. My conclusion after reviewing the record is that there is nothing sacrosanct about the original version of the example or a faithful telling of the underlying events - it is neither a watershed in Wikipedia policy history, nor a terribly elucidating example of Synthesis (as opposed to any other), nor a particularly well-written passage (the example or the text explaining it). If an example is useful for instructing readers new to the topic, convincing them to avoid synthesis, and serving as an analogy for policy debates elsewhere, then we should construct the example to best do that, and write it as well as we can. If the "true story" aspect to it helps the example be more convincing, fine. But we cannot be completely faithful to the original events without making the example murky, and picking up the underlying baggage of mud wrestling among partisan academics, questionable accounts, and Israel/Palestine Wiki-wars. Wikidemon (talk) 21:29, 3 March 2009 (UTC)

• I think I understand your reluctance. When you state Support or Oppose you can add a comment that states that you don't waive your right to redress in other ways. --Bob K31416 (talk) 21:47, 3 March 2009 (UTC)

We don't make policy using page surveys.

I can't even understand what the new version is saying. Whose "voice" is it? Did the article give the def of plagiarism, or is it the policy that's giving it? SlimVirgin ^{talk|contribs} 23:23, 3 March 2009 (UTC)

• SlimVirgin, Thank you for your comment. It would be helpful if you stated whether you Support or Oppose the proposed change. When you state Support or Oppose you can add a comment that qualifies your contribution to this survey, along the lines that you mentioned in your message. Thank you. --Bob K31416 (talk) 23:48, 3 March 2009 (UTC)

It should be clear from the indentation, format, and statement that the entire "example" is a hypothetical Wikipedia article, and thus in the voice of whichever hypothetical editors added the content. If some don't get that from the structure of the example, it may be better to improve the formatting or go to a yet simpler example. One thing that causes confusion is the case-within-a-case aspect, that example of poor sourcing we are putting forward is itself on the subject of sourcing. Wikidemon (talk) 23:39, 3 March 2009 (UTC)

It's a good example to use because it's a mistake that editors often make. In what way do you see the proposed version as an improvement? SlimVirgin ^{talk|contribs} 23:43, 3 March 2009 (UTC)

We could come up with an exact analogy on a topic other than sourcing. There are perhaps 12-15 ways in which the intermediate version in the current revert is convoluted, and the commentary is unduly informal and incorrect. I can list them all, but that would duplicate musch of the discussion from February 13-24 of the last archive.Wikidemon (talk) 00:00, 4 March 2009 (UTC)

Can you give a couple of examples of the commentary being unduly informal and incorrect? SlimVirgin ^{talk|contribs} 00:41, 4 March 2009 (UTC)

Okay (and with all due respect, because it appears you wrote most of this back in 2006): "Now comes the original synthesis:" -- "now comes" is informal and gives the article a narrative voice. In fact we are not telling this in chronological order so nothing is coming, and it did not happen "now". "Original synthesis" is not a real term. The word 'original' is awkward there. "The first paragraph was properly sourced" - that is not clear from the example; also we should be consistent about present versus past tense. "The second paragraph...expressed the editor's opinion that...Jones did not commit [plagiarism]." -- actually, it does not express that opinion, except by implication. "...a reliable source would be needed that specifically comments on the Smith and Jones dispute and makes the same point about the Harvard manual and plagiarism." - actually, that is not sufficient. The source would have to apply its point about the Harvard Manual to the Smith case and reach the same conclusion. Those are some of the things that could be tightened. I don't want to throw the baby out with the bathwater here, just try to keep things as clear as possible so that any reader who reads it should have an "a-ha" moment and understand very clearly what synthesis is and why it is wrong. Wikidemon (talk) 01:47, 4 March 2009 (UTC)

First, this is not my writing; as Kenosis said, it has been edited a lot. Your other points: (1) there is no need to be formal; this isn't the UN Security Council :-); (b) "now comes the ..." is a perfectly valid way to introduce the problem - it's not informal, and "now" doesn't refer to time; (c) original synthesis is as real a term as original research (and it wasn't me who added that); your point about what the source would need to say is, indeed, exactly what the version you're objecting to does say.

We can try to tighten it if you like, or revert to a clearer version, but the proposed version confuses the issue. SlimVirgin ^{talk|contribs} 01:55, 4 March 2009 (UTC)

I'm loathe to go through point by point all of the problems - my comments were correct - with the prior version. We improved it in many ways. Each change was an improvement. Better tone is better. Injecting a chronological account that didn't fit, and the inconsistencies with the verbe tense, were a problem. Original synthesis is not a defined term on Wikipedia. Original research is. The old version various editors complained about never claims in its example that Jones did not commit plagiarism yet the attached commentary incorrectly says it does. That's one of the various problems with it. Why hold out for a more poorly written, less precise version with tone problems? If there's a specific problem with the newer version let's figure out what that is and fix it.Wikidemon (talk) 02:49, 4 March 2009 (UTC)

The whole point of the example is the the WP article did claim that Jones hadn't committed plagiarism, according to definition X — a definition that was never mentioned by the sources but was merely the opinion of the editors. Your proposed version doesn't make clear who said what, and what was wrong with it. SlimVirgin ^{talk|contribs} 05:36, 4 March 2009 (UTC)

SlimVirgin, This discussion is not about the article that you mentioned, but it is about an example for WP:NOR. The example is not trying to be faithful to the article that it is based on and does not name that article, but is trying to give an example of a violation of WP:NOR. Any differences between the example and the article are for the purposes of clarifying WP:NOR, and it is not the purpose of the example to describe any specific article. I hope this has helped clarify the situation for you. Sorry for the interruption, please continue your discussion. --Bob K31416 (talk) 12:34, 4 March 2009 (UTC)

• I would change the final sentance "It appears that the conclusion was conceived by the editor, and is thus original research prohibited by this policy." to read something like: "Without providing a source that makes the conclusion, the final sentance is considered Original Research and not allowed in Wikipedia". But I believe overall the proposed wording is more clear and support the change.-- The Red Pen of Doom 00:55, 4 March 2009 (UTC)

Are there any remaining objections to replacing the example and discussion with the proposed version? --Bob K31416 (talk) 03:09, 5 March 2009 (UTC)

Yes, there are objections. SlimVirgin ^{talk|contribs} 03:24, 5 March 2009 (UTC)

Please state what objections of yours remain so it's clear what they are. Thank you. --Bob K31416 (talk) 03:31, 5 March 2009 (UTC)

The ones I was discussing above with Wikidemon. They remain unresolved, and I believe the discussion continues. :-) SlimVirgin ^{talk|contribs} 03:33, 5 March 2009 (UTC)

I am willing to carry on the discussion with you. Here is your last message in that regard. [20]

"The whole point of the example is the the WP article did claim that Jones hadn't committed plagiarism, according to definition X — a definition that was never mentioned by the sources but was merely the opinion of the editors. Your proposed version doesn't make clear who said what, and what was wrong with it."

Back then I made a comment as a bystander to the above comment of yours. I suspect that Wikidemon thought my comment was enough to end the line of discussion that you were following. Apparently it didn't. So let's pick it up here and I will repeat the main content of my message and await your response to my comment.

SlimVirgin, This discussion is not about the article that you mentioned, but it is about an example for WP:NOR. The example is not trying to be faithful to the article that it is based on and does not name that article, but is trying to give an example of a violation of WP:NOR. Any differences between the example and the article are for the purposes of clarifying WP:NOR, and it is not the purpose of the example to describe any specific article.

--Bob K31416 (talk) 04:11, 5 March 2009 (UTC)

The example is indeed trying to be faithful to the original, at least in broad terms. It would be easy to make up a SYN violation. It's not so easy to find a real example of one of the less obvious ways of SYN-ing. SlimVirgin ^{talk|contribs} 07:18, 6 March 2009 (UTC)

Providing an easily understandable example for novice editors is essential; providing an example that is representative of the more complex forms of syn is valuable in itself, but is secondary. --Phenylalanine (talk) 11:37, 6 March 2009 (UTC)

I agree with Phenylalanine that an understandable example for novice editors is essential. Both examples are, as SlimVirgin put it, trying to be faithful to the original, at least in broad terms. Both examples do not change policy. Both examples show editing that violates WP:NOR. So what is the important difference between the two examples? As indicated by a number of editors in this discussion, the proposed example is clearer.

The proposed example is clearly divided into three parts. 1) info from one source which is denoted by [27] 2) info from another source which is denoted by [28] and 3) the OR conclusion which combines the two and is denoted by bold font. The discussion afterwards simply and clearly gets to the point and states why there is a violation of WP:NOR.

The objection from one editor, SlimVirgin, had remained that the proposed example wasn't faithful to the original. That editor has since qualified that point with the opinion that the example should be faithful to the original in "broad terms". In any case the clarity of the example is far more important. The proposed example is clearer than the present example, as indicated by editor comments in this discussion. Even considerining the tenuous point of faithfulness to the original that one editor holds, which both examples have to some degree, it appears that there is sufficient consensus to replace the example because the proposed example is more clear. Thank you everyone for the discussion. --Bob K31416 (talk) 13:15, 6 March 2009 (UTC)

• Support - this is clearer. Note that the conclusion could be excluded on NPOV grounds as well since it is not patently obvious that references constitute "a source's information", and furthermore there is no discussion of other dictionaries' definitions of plagiarism and of the way such cases are handled in real life (e.g. case law, editorial policies, etc.) A more straightforward synthesis example would be one based on a simple logical deduction without interpretative and NPOV issues. --Phenylalanine (talk) 04:45, 5 March 2009 (UTC)

• Oppose. Second part of the example is supported by source but it is OR as Jayjg explained. Much better wording would be: According to a New York Times article on the controversy, plagiarism in this case is defined as using a source's information, ideas, words, or structure without citing them.^[28] -- Vision Thing -- 13:56, 6 March 2009 (UTC)

• Vision Thing, Thank you for your response, albeit after sufficient consensus was reached and the example was replaced. Although you wrote "Oppose" it appears that you are not opposing the replacement but are suggesting a change that is in neither of the examples. You are welcome to pursue that change in normal editing ways. I disagree with your modification because it is a needless complication to the example. Furthermore, as given by the example, the statement that you are suggesting is OR is not because it is supported by source [28], according to the premise of the example. But thank you for your thoughts anyway. --Bob K31416 (talk) 14:08, 6 March 2009 (UTC)

• Since SlimVirgin decided to revert my replacement of the example without any response on this talk page to my last message, either before or after the reversion, it looks like I have to post another message without waiting for that editor to respond on this talk page to my last one. And so it goes on.

SlimVirgin, you wrote in your edit summary, "rv the discussion is ongoing, and this anyway does not mention the issue of relatedness". Could you explain what you mean by "relatedness"? I looked at your previous remarks in this section and I couldn't find what you meant by "the issue of relatedness". Thanks. --Bob K31416 (talk) 17:51, 6 March 2009 (UTC)

• I oppose having a vote on this. Tom Harrison ^Talk 18:18, 6 March 2009 (UTC)

• It does seem silly to have a survey on a non-policy matter of cleaning up examples. This is more of a discussion than a !vote anyway, regardless of how the question was posed. For what it's worth I do kind of like Jayjg's version as reiterated by Vision Thing better than the one Bob K31416 and I came up with. Frankly, I don't really care what the subject matter of the example is... my hope was to simplify, streamline, and clarify it so that it more strongly reinforces the policy. If it's too murky it is still a good educational exercise to read and understand, but it is not as powerful a beacon to help explain and uphold the policy. Wikidemon (talk) 18:26, 6 March 2009 (UTC)
• Tom, Thanks for your concern, but this isn't a vote. Perhaps this quote from WP:Consensus will clarify the situation for you, "Polls are structured discussions, not votes. Opinion has more weight when you provide a rationale during a poll, not just a vote."     Note that almost all the editors have provided rationales.

BTW, I'm waiting for SlimVirgin's response to my last message and I even left a message on that editor's talk page informing of my response. --Bob K31416 (talk) 19:24, 6 March 2009 (UTC)

P.S. Tom, I just noticed that you are an administrator. So I'm puzzled by your remark, since you should already know all this about polls, votes, etc. ? --Bob K31416 (talk) 19:34, 6 March 2009 (UTC)

Bob, thanks for letting me know how the discussion is progressing. This page is on my watchlist, so I'll be automatically informed of any changes here from now on. If I have anything more to add I'll do so here - so please help me keep my thoughts organized by not posting on my talk page. Tom Harrison ^Talk 20:06, 6 March 2009 (UTC)

Not sure what to make of your remark, "so please help me keep my thoughts organized by not posting on my talk page" since I have never posted anything on your talk page. Could you clarify please? Thank you. --Bob K31416 (talk) 20:35, 6 March 2009 (UTC)

I didn't say you had; I just asked you not to, and explained why it is unnecessary (I'll see changes to this page on my watchlist) and undesirable (it will just confuse me). Tom Harrison ^Talk 20:49, 6 March 2009 (UTC)

Random break

Bob, the proposed version is wrong on a number of levels. First, it doesn't address the key issue of the source material needing to be directly related to the topic of the article. Secondly, it's actually wrong in identifying only the sentence in bold as OR/SYN; the previous sentence is OR too i.e. "Plagiarism is defined as using a source's information, ideas, words, or structure without citing them." It is OR because no source said this in relation to the dispute, which is the topic. The topic is not plagiarism. The topic is that particular dispute. Third, it leaves out the Harvard Manual as the source for the SYN (actually, it was the Chicago Manual of Style in the original; not sure why that was changed), and that matters because we need to show that, even though the source is a good one for plagiarism, it's not a good source for this dispute because it doesn't mention it. And fourth, it's quite hard to work out from the writing who is saying what. SlimVirgin ^{talk|contribs} 22:07, 6 March 2009 (UTC)

SlimVirgin raises a good point, we should clearly identify the first sentence as OR also. This would allow us to do two things at the same time — illustrate the syn rule as well as the "directly related" principle. --Phenylalanine (talk) 00:53, 7 March 2009 (UTC)

No, the first sentence is not OR at all, so calling it that would only confuse the issue. It is a statement that "Plagiarism is X" cited to a source that presumably says "Plagiarism is X". It is simply a second fact unrelated to the first, except by the following sentence, which is the editor's synthesis of the two. Of course the second sentence is inappropriate - the stage is set for synthesis when an editor mentions two things and draws a conclusion from the two that is not in the source. If the second fact had a sourced relationship with the first, then presumably the editor could have said that and sourced it, i.e. not synthesis. In that way synthesis usually involves inserting a new fact that is either not appropriate because it is unrelated. It's certainly possible for the editor to insert a second fact that can be or is sourced as relating to the first, then adds synthesis anyway. But whether the second fact is or is not related is an orthogonal issue to whether the editor's attempt to connect the two is synthesis. I have no objection to using a different example, but I don't think we should get hung up on finding the exact place synthesis occurs in an A[sourced]+B[sourced]=C[unsourced] construct. Wikidemon (talk) 09:35, 8 March 2009 (UTC)

"You must cite reliable sources that are directly related to the topic of the article." So, according to the policy, if a statement is relevant to the topic of the article, but is based on a source that is not directly related to the article topic, the statement is OR, whether or not the editor is engaging in original synthesis (i.e. advancing an original conclusion). This is fundamental. However, I tend to agree with you that the definition of plagiarism used in the Smith & Jones example is irrelevant and misleading, since the definition is ambiguous — do references count as a sources information? — and it can therefore be used to support both Smith's and Jone's contention. If we had a more precise definition of plagiarism in the example that showed without a doubt that Jones did or did not commit plagiarism, that definition would be truly relevant and so it would be original research within the context of the article if the source for that definition did not mention the Smith & Jones dispute. --Phenylalanine (talk) 02:29, 9 March 2009 (UTC)

<outdent>Here's a response to SlimVirgin's four points, point by point.

1) SV wrote, "First, it doesn't address the key issue of the source material needing to be directly related to the topic of the article." -

That is not the key issue of the WP:SYNTH section, as is clear from the part in WP:SYNTH that precedes the example.

"Do not put together information from multiple sources to reach a conclusion that is not stated explicitly by any of the sources."

"Editors should not make the mistake of thinking that if A is published by a reliable source, and B is published by a reliable source, then A and B can be joined together in an article to reach conclusion C. This would be a synthesis of published material that advances a new position, and that constitutes original research.^[1] "A and B, therefore C" is acceptable only if a reliable source has published the same argument in relation to the topic of the article."

The key issue in WP:Synth is obviously the issue of putting together info from sources A and B to reach a conclusion C that is not stated in the sources. If source material is not directly related to the topic of the article, that is an indication that it is going to be used to reach a conclusion that is OR. The sourced material itself is not the OR. The conclusion is the OR.

2) SV wrote, "Secondly, it's actually wrong in identifying only the sentence in bold as OR/SYN; the previous sentence is OR too i.e. 'Plagiarism is defined as using a source's information, ideas, words, or structure without citing them.' It is OR because no source said this in relation to the dispute, which is the topic. The topic is not plagiarism. The topic is that particular dispute." -

That was answered above in item 1. I should add that SlimVirgin did not properly quote the proposed example because an important part of the sentence was left out, i.e. the citation [28]. Here's the whole sentence.

Plagiarism is defined as using a source's information, ideas, words, or structure without citing them.^[28]

This sentence itself is not OR because it is sourced. --Bob K31416 (talk) 03:02, 7 March 2009 (UTC)

That first sentence is OR. The source doesn't mention the particular dispute, so what we have here is a Wikipedian casting around for a definition that no source has presented in relation to the topic, then using that definition to decide that Jones did commit plagiarism, when in fact the issue remains open, even if you accept the validity of the source. One side says using a book's sources without acknowledgement is plagiarism; the other side says it's best not to do it, but it's not plagiarism, whereupon a Wikipedian arrives to settle the matter with a self-selected definition. Classic SYN. Posting the definition and the conclusion derived from it are both examples of OR. SlimVirgin ^{talk|contribs} 05:25, 7 March 2009 (UTC)

SV, Syn means using sourced info A and sourced info B to conclude C, which is not sourced. Just as this isn't clearly addressed in your above comments, so too it isn't clearly addressed in your version of the example. And as I mentioned before, the first sentence by itself is not OR. Together with the other sourced part that preceded it, they are both part of the OR process that leads to the unsourced conclusion which is the OR. --Bob K31416 (talk) 10:51, 7 March 2009 (UTC)

You know, I think the whole thing is a matter of definition. We can all agree that it is wrong but seem to disagree on where it went wrong. That's a bit angels-on-pinheads, like asking exactly where the mistake occurs in 3+5=7.Wikidemon (talk) 09:41, 8 March 2009 (UTC)

3) SV wrote, "Third, it leaves out the Harvard Manual as the source for the SYN (actually, it was the Chicago Manual of Style in the original; not sure why that was changed), and that matters because we need to show that, even though the source is a good one for plagiarism, it's not a good source for this dispute because it doesn't mention it." -

In the proposed example, the statement is said to be properly sourced. Stating a specific source that the reader may or may not recognize as an authority does nothing to clarify the situation. Please note that the editors that are reading WP:SYNTH for guidance are from all subject areas of Wikipedia, from puppies to plutonium, and can't be expected to know anything about either the Chicago or Harvard manuals. Also, as I mentioned in item 1)

The key issue in WP:Synth is obviously the issue of putting together info from sources A and B to reach a conclusion C that is not stated in the sources. If source material is not directly related to the topic of the article, that is an indication that it is going to be used to reach a conclusion that is OR. The sourced material itself is not the OR. The conclusion is the OR.

And here's something from the lede of Wp:NOR.

To demonstrate that you are not presenting original research, you must cite reliable sources that are directly related to the topic of the article, and that directly support the information as it is presented.

This doesn't say that the unrelated info from sources are themselves the original research. The unrelated info is part of the A+B that yields the C which is OR. It doesn't say that A and B individually by themselves are each OR. Also note that if A is related to the topic and B is not, they are both part of the OR process to reach the OR conclusion C.

4) SV wrote, "And fourth, it's quite hard to work out from the writing who is saying what. " -

I don't understand the above remark of yours at all. The following is from the proposed example.

Smith claimed that Jones committed plagiarism by copying references from another author's book. Jones responded that it is not plagiarism, but acceptable scholarly practice to copy references from other people's books.^[27]

As indicated, these are the remarks of Smith and Jones that were reported in reference [27].

Furthermore

SV, Your example seems to be missing something important. It doesn't clearly show what A and B is in the relation A + B = C which is what the WP:Synth section is all about. The proposed example clearly shows this. --Bob K31416 (talk) 03:02, 7 March 2009 (UTC)

• Weak oppose. The new example illustrates OR of the type where a statement is made which is a Wikipedian's conclusion not directly expressed by a source. But the original example illustrates something else: OR of the type where the Wikipedian article does not make any unverifiable statement, but places verifiable statements beside each other specifically in order to imply a particular conclusion. It's important to illustrate this. The wording around the original example could be improved (added to) to explain it better. I marked my oppose as weak because I haven't read the whole discussion (sorry). ☺Coppertwig (talk) 01:52, 17 March 2009 (UTC)

To make an observation, the example is based on a real case, involving two academics in a controversial debate on Israel/Palestine. As it is such a controversial area, and there are real people with real accusations of plagiarism and real reputations on the line, I suggest that we at Wikipedia take care to ensure that we fairly represent the original case if we are to use this example.

At the moment, I think the example doesn't bear much resemblance to the original case in certain areas. For example, in the second part of the original case, the wording under scrutiny was broadly as follows:

"...plagiarism is defined as using a source's information, ideas, words, or structure without citing them._{Source:Harvard manual} Smith suggests that this copying of quotations amounts to copying ideas _{Source: Smith specifically saying that Jones is copying ideas, specifically cross referencing the Harvard manual definition....}".

At the moment, I don't think the wording in the example reflects the reality of the case very well (indeed, while they are not very well worded, I don't think the above sentences are a particularly good example of OR - the Harvard manual definition was clearly mentioned in a source related to the article). Enchanter (talk) 01:19, 25 March 2009 (UTC)

I don't know if you noticed but the discussion continues quite some distance below. Adding the writers' names to the policy page raises BLP concerns. It would be distracting given that one very famous person is involved and it is on the subject of the Israel-Palestinian dispute, and the on-wiki part involves edit warring and an SPA who appears to be a sock. Plus, the real example, though clearly synthesis, is not as clear-cut as the abstracted examples we are talking about. If it were a choice between being good Wikipedia historians and making up an example, I would side with making up an example. Wikidemon (talk) 03:46, 25 March 2009 (UTC)

A slight modification of proposed version to eliminate issue of alleged OR sentence

The following modification of the first sentence of the second paragraph of the proposed example should eliminate the issue of alleged OR that has been brought up. Although I feel the previous version was correct, in order to move the process of acceptance along more quickly, I am willing to make this slight modification. Essentially the change is an additional phrase, "In the dispute", at the beginning of the sentence.

Change from:

Plagiarism is defined as using a source's information, ideas, words, or structure without citing them.^[28]

To:

In the dispute, plagiarism was defined as using a source's information, ideas, words, or structure without citing them.^[28]

And as mentioned in the proposed version, this statement is properly sourced by reference [28], which is a premise of the example. Everyone is invited to comment. --Bob K31416 (talk) 11:21, 8 March 2009 (UTC)

--Bob K31416 (talk) 00:37, 9 March 2009 (UTC)

I reread the discussion and noticed that the above modification has the same essential feature that was first proposed by Jayjg, and a similar modification was later proposed by Vision Thing and supported by Wikidemon.

From Jayjg:

According to a New York Times article on the controversy, plagiarism in this case is defined as "xyz".[28]

From Vision Thing:

According to a New York Times article on the controversy, plagiarism in this case is defined as using a source's information, ideas, words, or structure without citing them.^[28]

Although I don't think that mention of the New York Times is necessary, since reference [28] takes care of any sourcing, I do see some better wording in the above suggestions by Jayjg and Vision Thing. So I would suggest the following modification instead of the one that I first mentioned in this section. This modification takes the last part of Jayjg's and Vision Thing's suggestions.

Plagiarism in this case was defined as using a source's information, ideas, words, or structure without citing them.^[28]

In this way, the definition of plagiarism and the dispute are connected together by reference [28], which was the point of all three suggestions mentioned above. With this modification, it appears that the last remaining objection to the proposed version has been satisfied. Here is the proposed version with the current slight modification discussed above.

--Bob K31416 (talk) 20:36, 9 March 2009 (UTC)

• Support I like the change. The example is concise and straightforward. --Phenylalanine (talk) 22:37, 9 March 2009 (UTC)
• As explained already, the example is inappropriate since the sentence supported by 28 is a SYNTHesis. There's no indication that source 28 is an appropriate definition of plagiarism in this case. In order to know which source is appropriate, we need one that discusses plagiarism in the context of the specific discussion. And no, I'm not going to vote here. This isn't a vote, it's a discussion. Jayjg ^(talk) 21:31, 10 March 2009 (UTC)
  • Bob, there's obviously no consensus for your new version, which is inferior to both the version above and below. Please don't re-insert it. Jayjg ^(talk) 02:42, 24 March 2009 (UTC)

Jayjg ^(talk) 03:14, 13 March 2009 (UTC)

If source [28] directly supports the sentence "Plagiarism in this case was defined as using a source's information, ideas, words, or structure without citing them." as it is presented, then [28] is necessarily directly related to the Smith & Jones dispute. Jayjg, your example would work if you said instead "Plagiarism is defined as using a source's information, ideas, words, or structure without citing them." Apart from that, I like your version. --Phenylalanine (talk) 23:12, 10 March 2009 (UTC)

Phenylalanine, Like you and I have mentioned, now the sentence is directly related to the Smith & Jones dispute. I discuss this some more in my message below to Jayjg. --Bob K31416 (talk) 02:12, 11 March 2009 (UTC)

Jayjg, re your remark, "There's no indication that source 28 is an appropriate definition of plagiarism in this case." -

Source [28] is not defining plagiarism. It is only describing what was defined in the Smith/Jones dispute. Please note the beginning of the sentence:

"Plagiarism in this case was defined ..."

Note that the sentence said it "was defined" which is different than the previous version of yours which said it "is defined". The "was" makes all the difference since it establishes that source [28] is reporting that in the dispute, plagiarism was defined, rather than [28] defining plagiarism itself. I note that in your most recent version above, you use "was". Perhaps you weren't aware of the change from "is" to "was" when you copied the version? --Bob K31416 (talk) 02:12, 11 March 2009 (UTC)

You're right, I read "is" rather than "was", and didn't realize the change. I've modified my version so that it's closer to the original, and deals with two OR issues rather than just one. Jayjg ^(talk) 03:13, 13 March 2009 (UTC)

Thanks Jayjg. It appears that you now recognize that the subject sentence in "Proposed version with the above modification", is not OR.

Here is the sentence, just so there's no misunderstanding as to what we are discussing:

Plagiarism in this case was defined as using a source's information, ideas, words, or structure without citing them.^[28]

Do you agree that this sentence is not OR? Thank you. --Bob K31416 (talk) 17:05, 13 March 2009 (UTC)

Where are we on this? Wikidemon (talk) 17:15, 15 March 2009 (UTC)

Waiting for Jayjg to respond. I think Jayjg takes off from sundown friday to saturday night for religious reasons, so I'm trying to be considerate of Jayjg's religious practices. However, I have noticed that Jayjg has been active elsewhere on the Wikipedia since saturday night. --Bob K31416 (talk) 19:30, 15 March 2009 (UTC)

The example, as I've given, is more informative - it describes both types of OR, both using a source not directly related to the topic of the article, and the drawing of a conclusion not found in the sources. That is also the strength of the original version, though my amended version states it more clearly. Jayjg ^(talk) 19:59, 15 March 2009 (UTC)

Please respond to my previous question so it is clear whether or not you agree that the sentence with the slight modification is no longer OR. Thank you. --Bob K31416 (talk) 20:07, 15 March 2009 (UTC)

I've already given all relevant responses. Please respond to the point that your proposed re-write only addresses one form of OR, rather than two, and is therefore not preferred. Jayjg ^(talk) 20:12, 15 March 2009 (UTC)

<outdent>Jayjg, It's not clear why you don't want to clearly say whether or not you think the subject sentence is OR. In any case, you've convinced me that you will not cooperate on this, so I won't pursue it.

However, this won't preclude me from cooperating with you and responding to your request, which was, "Please respond to the point that your proposed re-write only addresses one form of OR, rather than two, and is therefore not preferred." First, to give credit where credit is due. The re-write that I prefer was the result of discussions with other editors in a previous section and in this section. And one of the editors that influenced the re-write was you, as I indicated at the beginning of this sub-section.

Both re-writes, the one that you prefer and the one that I prefer, are proposed to go into the WP:SYNTH section. The re-write that I prefer gives an example of a violation of WP:SYNTH. If I understand you correctly, the re-write that you prefer gives an example of a violation of WP:SYNTH that includes an example of a violation of the directly related principle that is discussed in the Reliable sources section.

I prefer the example that gives a simpler description that involves one form of OR because it is easier to understand by an editor who is looking for a clear example of a WP:SYNTH violation. --Bob K31416 (talk) 05:20, 16 March 2009 (UTC)

Of the three possibilities on the page, the first and third are superior to your proposal, since they both highlight two common WP:SYNTH issues, rather than just one, as your example does. Both are issues of WP:SYNTH, not one of SYNTH and one of a "directly related principle". I believe the third is better than the first, since it states more clearly what both problems are, but clearly either is better than an example which doesn't even address the issue. Jayjg ^(talk) 04:20, 17 March 2009 (UTC)

I looked and looked but I couldn't find anything of substance in your remarks to respond to! You are completely ignoring the issue of CLARITY. Editors seeking help with NOR don't understand the present example nor will they understand the example that you are proposing because it is unclear since it tries to do more, and in an unclear way, than simply show the basic violation of WP:SYNTH. Frankly, if there isn't a change in your attitude, I don't think any further discussion with you is worthwhile. You're too obstructive. BTW I saw how you and SlimVirgin got the infamous example into the Synth section by edit warring.[21] That is apparently both of yours' style. You two are quite a team, much to the detriment of WP:NOR.--Bob K31416 (talk) 05:52, 17 March 2009 (UTC)

The third option is obviously clearer than the second, since it addresses multiple NOR issues, and there's no evidence editors don't understand the current example. What is unclear is why you would prefer a less informative example to a more informative one. Regarding the rest, please restrict your comments to a discussion of policy, rather than other editors. Jayjg ^(talk) 00:26, 20 March 2009 (UTC)

You're amazing. Re "...there's no evidence editors don't understand the current example." You know this "of course", so this is for other editors to read, who may be following this discussion. You're beyond help. This looks like the end of my discussion with you. It's just not worth my time. --Bob K31416 (talk) 04:51, 20 March 2009 (UTC)

Hey, folks. Sorry to have ignored this for a few days, I got waylaid by yet another Obama debacle. I think we're getting lost in the details and perhaps worn down by frustration. I appreciate BobK's tenacity for working through this, but with all due respect Jayjg's proposal is also a very good one. I find it just as hard to understand as the current version (it took me a minute or two to read slowly and puzzle it out), but for any editor who takes the time there is a good reward. It's dense with policy wisdom, not logical knots that have to be untied. Whether we characterize the example as illustrating two acts of synthesis, or as one predicate act of a relevancy mistake plus one act of synthesis, it is the same thing. There are three steps to synthesis: (1) introducing a first fact that is (one hopes) pertinent and appropriate to the article; (2) introducing a second fact that is not sourced as pertinent to the first fact; and then (3) making an unsourced conclusion that combines the two facts. An example that illustrates only the final act of synthesis would be easier to read, but correspondingly less informative. I'm fine either way because either would be an improvement.Wikidemon (talk) 15:10, 22 March 2009 (UTC)

Re your remark about both Jayjg's proposal and the current version of SlimVirgin, "I find it just as hard to understand as the current version". That's the problem. If it's hard for you, who have had extensive experience with it on this Talk page, try to understand how hard it would be for an editor coming to WP:NOR for help. --Bob K31416 (talk) 12:03, 23 March 2009 (UTC)

I hear you, but maybe it's just a complicated subject. I'd rate myself mid-level in terms of understanding citations and verifiability, not one of our experts. You know how hard it is to explain to people, and we're even having some trouble coming to grips with what it means, not just how to describe it. Maybe the main explanation can be simple, and the example is an elaboration that serves as a tutorial and reference point. Also, perhaps someone good at illustration could come up with a graphic to illustrate why A + B does not equal C. Wikidemon (talk) 16:10, 23 March 2009 (UTC)

NOR is indeed a complicated subject, which is why it requires strong examples that really tackle it, rather than weak ones that avoid it. I'm fine with the longstanding version, or the improved version I've suggested. I do not, however, support Bob's weakened version; that would be a step backward. Jayjg ^(talk) 02:42, 24 March 2009 (UTC)

I made a mistake in reverting SlimVirgin because I didn't quite see what was going on. I think we're all set now. I am fine with either of the new proposed versions, and I favor them both over the current version. I think if we set up a decision matrix, then as long as Bob K31416 doesn't like the new version less then the current version, that means everyone is either equally pleased or more pleased now than before - which is as good as we can expect. As for it not being the best version in everyone's eyes, maybe apply that aphorism about not letting the perfect be the enemy of the good.Wikidemon (talk) 02:55, 24 March 2009 (UTC)

Comparison of versions

There is considerable support for the new version that I included in WP:NOR because it is clear. The old version of SlimVirgin, the next version of SlimVirgin, the proposed version of Jayjg all have the same problem of being unclear because they try to include a "directly related" aspect. Jayjg knows that it is unclear from personal experience.[22]. Not only is it unclear but it is misleading. It is misleading because it implies that violation of WP:SYNTH requires that one of the elements must not be directly related, which is false.

Currently SlimVirgin is conducting an edit war to keep out the new version. This is similar to the edit war that SlimVirgin and Jayjg used to get their example into WP:NOR in the first place.[23]. SlimVirgin has made 3 reversions in the last 10 hours. Also note that SlimVirgin hasn't made a comment in this section since March 7, 2009, 17 days ago, in the finest tradition of an edit warrior. At present, the version that SlimVirgin wants is in the article. I have not made any reverts since SlimVirgin's last revert. --Bob K31416 (talk) 11:56, 24 March 2009 (UTC)

You're lowering the bar for "considerable support" to a level I'm not comfortable with on policy discussion. I may be alone on this one, but giving the discussion a quick once-over, I actually prefer the original example (the one that concludes that plagiarism was not committed) because it's more complex — it does a better job of teasing out what makes synthesis synthesis. Reading the new example just makes me think "duh" and I doubt that anyone serious about encyclopedic writing would ever think that was an appropriate sentence to write. The original example provides a much more real-world situation on Wikipedia that is come across every single day, written by well-meaning authors. I acknowledge the potential problem of the implication that synthesis requires an unrelated source, but that's not anything that a quick note can't make abundantly clear. —bbatsell ¿? ✍ 14:51, 24 March 2009 (UTC)

I think this amounts to aguing over "how many angels can dance on the head of a pin". Both versions convey the idea that sythesis is bad. One (the "long standing" version) is overly complex and confusing, the other (Bob's new version) is overly simplified. I really think we need to jettison the entire plagerism example and come up with something that hits a middle ground... relfecting a typical example of synthesis that even newbees can understand. Blueboar (talk) 16:00, 24 March 2009 (UTC)

I don't mean to be intemperate, but if someone finds the original example too complex and confusing to understand, they probably shouldn't be editing an English-language encyclopedia. Regards, —bbatsell ¿? ✍ 16:09, 24 March 2009 (UTC)

I am sure that we have all faced situations where we would agree with that sentiment.... unfortunately we are stuck with "anyone can edit"... and that means anyone, even those who probably shouldn't. We need to write policy so that anyone can understand it. Blueboar (talk) 16:17, 24 March 2009 (UTC)

Part of the problem is that much of the complexity is a red herring: the original example contains several errors and ambiguities that are not specifically synthesis, e.g. the phrasing of an impertinent issue in conditional tense ("If Jones did not consult the original sources..." - consulting is not the issue, citing is). Another part of the problem is that the explanatory commentary is written in an informal, non-authoritative tone, and is vague and possibly inaccurate to boot. It is a shame that a simple clean-up requires such hand-wringing. With all due respect, until the arrival of a newer batch of editors who missed the now month-long discussion, Bob K31416 was the only one hanging out for the simplified example. Everyone else either favored, or was neutral on, the more complex version that preserves the heart of the original by illustrating same type of three-step synthesis: (A) cited fact + (B) cited but not directly related fact --> (C) uncited conclusion. My last comment to Bob before signing out last night is that if he finds the new example better than the current one, everyone else is either neutral or prefers it as well, so at least we will have made an improvement. Bob's preferred version is strongly considered by at least two editors here (and now more) to be weaker than the original, so I do not think it can gain consensus. It is true that there are other flavors of synthesis as well, but this three-step synthesis is a common mistake in articles, and a pattern people should be watching for. Wikidemon (talk) 17:29, 24 March 2009 (UTC)

An irrelevant fact is a predicate to some kinds of synthesis. But adding the fact or not is a relevancy question, not original research. A fact verified to a reliable source is, by definition, not original research. We need not reach that issue here, though. Wikidemon (talk) 18:12, 24 March 2009 (UTC)

But mixing of facts, each individually verified to a reliable source, can indeed be original research - the policy talks about this. A great deal of OR mischief is done when cited, though not directly related, material is added to articles. Jayjg ^(talk) 23:44, 24 March 2009 (UTC)

Bob, you claim there is "considerable support for the new version" you inserted. Can you quantify that please? Exactly who supports it? Jayjg ^(talk) 23:44, 24 March 2009 (UTC)