Wikipedia talk:No original research

Citing oneself

I'd like to remove this section, which is a little outdated:

If an editor has published the results of his or her research in a reliable publication, the editor may cite that source while writing in the third person and complying with our neutrality and conflict of interest policies. If you are able to discover something new, Wikipedia is not the place to premiere such a discovery. This policy does not prohibit editors with specialist knowledge from adding their knowledge to Wikipedia. In fact, expert input is encouraged and experts often have specific knowledge of the relevant literature. However, as with all editors, this policy does prohibit experts from drawing on their personal knowledge without citing reliable sources. See also Wikipedia's guidelines on conflict of interest.

We could replace it with something like:

Citing your own published work is discouraged as a conflict of interest. If you have written material that has been published by a reliable, independent publisher, by all means suggest it for inclusion on the article's talk page.

Any thoughts?

SlimVirgin ^{talk|contribs} 23:37, 2 January 2011 (UTC)

In practice, there probably are a lot of conflicts of interest when people cite themselves. However, saying that we discourage people from citing themselves may be perceived as hostility towards those who actuallly know what they are writing about. I challenge you to find any scholarly journal that discourages authors from citing themselves. I fear that the statement in its present form could be viewed in isolation to portray Wikipedia as anti-intellectual. Jc3s5h (talk) 01:07, 3 January 2011 (UTC)

Scholarly journals have checks and balances that we don't have. And the analogy doesn't really hold because scholarly journals don't require neutrality, or that works cited be the most notable. In addition, writers cite themselves in their own articles, something that never applies to an article on Wikipedia.

The clause encouraging people to cite their own work is (as I recall) a few years old, and harks back to before COI became a big problem for us. If someone's work is worth citing, someone else will do it. SlimVirgin ^{talk|contribs} 05:38, 3 January 2011 (UTC)

Not necessarily - it depends how much attention anyone else might be paying. I don't see anything wrong with citing a reliable source which was written by oneself - edits should be judged on their merits, not by who did them - though we could say that anyone doing this is encouraged to leave a note on the talk page saying that they've done it. --Kotniski (talk) 09:42, 3 January 2011 (UTC)

What's wrong with it is that they're more likely to cite themselves than anyone else is, and more likely to decide that their paper is more important than anyone else's. Classic COI. This is one of the many ways WP is used to self-promote. SlimVirgin ^{talk|contribs} 20:42, 3 January 2011 (UTC)

If an editor uses a pseudonym here, and wishes to cite a paper she wrote, is she supposed to write on the talk page "My real name is Joan J. and I wrote this paper..."? That would be absurd, and even asking the pseudonymous editor whether she wrote the source could be construed as a violation of the OUTING policy. So that seems like a bad idea.

I generally agree with Kotniski that we should judge sources by their apparent quality, not by trying to divine whether the editor who added them also wrote them. — Carl (CBM · talk) 12:50, 3 January 2011 (UTC)

Agree. The proposed change would often require people to out themselves. Also this would also impair experts from writing. For example, having written the definitive source would prohibit Einstein from writing about the theory of relativity in Wikipedia. Could not write without citing, and could not cite the definitive source. North8000 (talk) 13:08, 3 January 2011 (UTC)

But someone else would do it if the paper were important. SlimVirgin ^{talk|contribs} 20:42, 3 January 2011 (UTC)

Still, we can't write a policy that encourages people to out themselves. Given our anonymity policies, we can only treat COI by the duck test anyway. If someone is adding material in a way that appears to unduly promote an author or viewpoint, we have ways to deal with it that don't care whether the person is the author of the material.

Also, there's no requirement that a paper has to be "important" to be included in an article. If someyone (author or not) adds a paper as a reference to an article where the paper is clearly directly related, that's not a policy violation. If someone adds a paper to 15 articles where it is unrelated or barely related, other editors will notice and handle it. — Carl (CBM · talk) 21:28, 3 January 2011 (UTC)

Hello Slim. I used Einstein as an extreme example, but the unintended consequence is that prohibiting citing could prohibit writing. A more realistic example is that the only way a prominent expert could (cited) write would be by citing their rival's works, which they probably wouldn't do. I know of cases where this happened, despite there being no prohibition against it, because they didn't want to risk COI accusations for the privilege of writing for free in Wikipedia. North8000 (talk)

We don't want experts here as experts, but as Wikipedians, and as such they must cite their rival's work if it's appropriate. But bear in mind that we're largely talking about unknown academics and writers here, because known people will be cited by others. What this clause does is encourage non-notable writers to use Wikipedia to promote their own work, and that's why I think we ought to get rid of it. SlimVirgin ^{talk|contribs} 21:53, 3 January 2011 (UTC)

When including a reference in a Wikipedia article, it doesn't matter much if the reference is cited by others or not. References published by academic presses and professional journals are presumed reliable regardless whether they are cited anywhere at all.

The language above doesn't really encourage anything. The conflict of interest policy says, "Editing in an area in which you have professional or academic expertise is not, in itself, a conflict of interest. Using material you yourself have written or published is allowed within reason, but only if it is relevant and conforms to the content policies." That seems perfectly reasonable to me, and the language above agrees with it. I don't see that anything encourages excessive self-citation.

Moreover, if the NOR policy were to try discouragement of all self-citation, how could that possibly be enforceable in circumstances when the present language isn't? — Carl (CBM · talk) 22:10, 3 January 2011 (UTC)

I'll point out that I have actually been somewhat active lately in dealing with COI related to a banned user. In that situation, as is often the case, there is no way to tell who is actually adding the material. So it doesn't help to have a policy that directly bans self-publication; the key policy is NPOV. In the case I've been involved with, there is also an arbcom case that allows for stricter action that we would normally take. — Carl (CBM · talk) 22:17, 3 January 2011 (UTC)

The point should surely be that other people decide whether to cite your work (and we don't just mean academic work; we talk only about "reliable publication"), just as other people should decide whether to create an article on someone. I take your point that we often don't know who is adding material, but we often do. The point is simply not to encourage it, enforceable or not.

As to whether we encourage it, we say: "If an editor has published the results of his or her research in a reliable publication, the editor may cite that source ..." To me, that says "go right ahead." SlimVirgin ^{talk|contribs} 22:33, 3 January 2011 (UTC)

Hello Slim. I digress, but responding on your first point, I think that it's one of the least noticed and most denied realities in Wikipedia that a whole lot of the good stuff is written first by people who know the topic and THEN its cited, including by using the expertise to select good sourcing and leave out wrong or misleading stuff from sources. I guess people's opinions here are molded by what issues they've most run across. I see contentious articles that are eternal failures because intelligent accurate writing can be kept out, (and so am sensitive to the issue of further discouraging sourcable intelligent writing) and you are seeing more situations of marginal or bad source insertions due to COI's. I think we just wander different parts of WP. :-) Sincerely, North8000 (talk) 22:41, 3 January 2011 (UTC)

I think it is better to write up some guidelines aimed at all editors to help them spot problematic cases of excessive self-citations. Suppose professor X comes here to start an article on theory Y and there are a lot of citations to his/her own work. Our attitude should be that it is good to have that article and Professor X contributing here, but we also want to make sure that all these citations are really appropriate. Count Iblis (talk) 01:18, 4 January 2011 (UTC)

That's the real key here... if the COI editor is acting appropriately, following Wikipedia's rules, and contributing good material, I don't think it really matters if that material happens to include his/her own work. If they are acting inappropriately then there is a problem. We want to encourage the former, and discourage the latter. I think our current language does this. Blueboar (talk) 03:52, 4 January 2011 (UTC)

Agree with Blueboar, I think the current version is superior. We have to tread a fine line here. On the one hand, we want to discourage dubious editing when there is COI, but on the other hand we want to encourage editing by people who are experts in the field, who if they are academics, would likely have published some papers in the field. I think the current version balances this nicely. The proposed version would would unnecessarily scare away academics who have published in that field. LK (talk) 04:34, 4 January 2011 (UTC)

Without pretending to offer an opinion on the word change, I'll offer a few perspectives on the issue in general. I wonder if we should be committal on the question at all in the NOR policy--because so far I don't understand how it is a quandary here in terms of NOR. And because even authors of scientific papers legitimately and routinely cite their own previously published work as reference when justified, I wonder if we've clearly defined a valid rationale to prohibit the practice on wiki. Off-wiki conflicts-of-interest are mostly focused on transparency with "source of funding" (not citing oneself) because issues like citing one's own work has an overt transparency--authors of the work, like the sources cited, are identified by name. But wp invites anonymous editing, and editing-for-pay is difficult to impossible to detect (ie non-transparent) to readers OR fellow editors. So predictably here rather than enforcing "transparency" the policy against citing one's own work may act as a gotcha against real life authors who have disclosed their identities and have valid, notable, published, citable opinions but don't comprehend that an "anybody can edit" website might raise a fuss about it. I've seen a half-dozen cases where this has happened-where new editors are roughed-up on wiki for referencing claims that are everywhere else (including peer-reviewed publishing) perfectly justifiable and cited to something they've authored. So if the "anyone can edit" wiki has unique transparency issues to deal with maybe we shouldn't misidentify or dismiss them out-of-hand as COI violations and instead alert legit authors who venture into the odd wiki-world with good intentions to free-up good content under CC-BY-SA 3.0/GFDL. And for the legitimate self-citations (and I've seen many hostilely objected to as COI transgressions) maybe a more diplomatic guideline (ie "caution to experts") is called for. COI is too loaded a term to blanket across the board to cover the spamming with legitimate self-citations in one broad swath. Professor marginalia (talk) 07:19, 4 January 2011 (UTC)

I think that the WP:COI guideline does a good job of defining a REAL COI, and covers things like this. Roughly speaking, it's when self interests are allowed to override WP objectives. But, like many WP policies/guidelines, it is routinely mis-stated. Even in this discussion 1-2 folks have inadvertently done that, by using "COI" as shorthand for any situation where an editor cites something that they wrote. My focus isn't about the citing per se, it's about the consequences.....officially, if you don't cite you don't write. North8000 (talk) 12:57, 4 January 2011 (UTC)

From a purely NOR perspective, the most important line in the paragraph is: "If you are able to discover something new, Wikipedia is not the place to premiere such a discovery". We definitely don't want to lose that. Blueboar (talk) 13:15, 4 January 2011 (UTC)

Suggest keeping that sentence, putting it back where it came from, and deleting the rest of the section. --Bob K31416 (talk) 01:40, 5 January 2011 (UTC)

The paragraph is a useful explanation of when it's appropriate and when it's inappropriate to reference research done by oneself when editing an article. Why should it be removed? LK (talk) 10:52, 6 January 2011 (UTC)

I've moved the "most important line in the paragraph" to the beginning of the paragraph to highlight it. This also makes the paragraph flow better and more logically. LK (talk) 10:59, 6 January 2011 (UTC)

Seems like a digression into citing one's own published material. Also note that the second sentence contradicts the first, which is not good flow. Have you considered putting the topic of citing oneself in WP:V? --Bob K31416 (talk) 13:58, 7 January 2011 (UTC)

The proposal seems sort of misguided and sort of absurd. There has to be a vast area of as-yet-unknown knowledge for which, once known, there is good reason to have it in Wikipedia - once it has been published elsewhere. The misguided aspect is the apparent assumption that the original discover or developer of an idea has a selfish, incompatible-with-Wikipedia nature and motivation. This amounts to an attitude that those who are devoted to finding and publishing new things are unqualified to write Wikipedia articles about the new things, which seems to be just about exactly 180 degrees in the wrong direction. One of the virtues of Wikipedia is that it is capable of rapid improvement, rapid coverage of useful facts. (When some person of note dies the date of death almost instantly appears in Wikipedia.) That is a valuable attribute.

I do not know where I could find a reference to justify saying in Wikipedia (were it worth saying, which I think it is not) that if you are using a garden hose and wearing shoes and turned the hose so that it pointed downward your shoes would likely get wet. So (again, if it were worthy of inclusion) I could do the experiment, publish the shoes-get-wet observation somewhere respectable (a step that is hardly necessary other than to comply with the letter of the policy), and then include it, wherever appropriate (were it so), in Wikipedia. The proposed policy would have it that I am a suspect source for such material. It is fully verifiable, it is added in good faith, it is not in any way banging the drum for any agenda, but the proposed policy wording would have it be treated as suspect - and subject to deletion by some over-eager editor. Or I might develop/discover something useful and notable in fighting spam, or develop simple formulas for computing something (such as Savitzky-Golay smoothing coefficients) that was hitherto unknown.

I think policy writers would be well advised to stop and think about whether the policy about to be proposed does a disservice to both Wikipedia and the capable editors who are willing to share what they know -including what they know that they have themselves developed and published - by creating hurdles aimed at the few who have suspect motivations but which hamper all. Minasbeede (talk) 00:14, 16 January 2011 (UTC)

Here's another problem with the suggested change. the suggested change says self-citation is "discouraged." Not forbidden, just "discouraged." Imagine a situation in which a reasonable, well-meaning editor does an edit in which the editor cites the editor's own work after first determining that doing so is not an instance of "conflict of interest." But then a literal-minded editor comes along, sees it is a self-citation, and deletes it because of the policy. the policy doesn't forbid, it merely discourages. The literal-minded editor cares not for any such nuance, the literal-minded editor is interested almost entirely in finding instances of violation of policy and then removing the offending material - even if it does not violate policy but may be accused of doing so because of the imprecision of the policy language. If the first editor cares enough about this to object, to re-inset the material and then open up a discussion on the talk page (or just open up the discussion) the question will probably not be determined at all on the merits of the first editor's position but on amateur lawyering over what the policy says - or over what it can be bent to say. The offense, if an offense has occurred, is "conflict of interest." No doubt that can occur, but when it does the reason for removing text is "conflict of interest," not self-citation. The proposed policy change would make matters worse by creating yet another tool that can be abused by the nit-picky.

Yes, I recognize that the first editor might self-deceive in the editor's determination that the self-citation is not a conflict of interest. The point just made was that the offense is the conflict of interest and that conflict of interest is sufficient reason to remove the edit. That an editor may so err does not blacken all self-citation, as indeed the proposed policy wording appears to recognize. The objection to the proposed wording is that it creates a new problem and solves no problem. Minasbeede (talk) 03:41, 16 January 2011 (UTC)

"If someone's work is worth citing, someone else will do it." Sadly this assumption is not correct. The majority of pages sit for year after year as stubs, without anyone really updating them. The idea that if it's important someone will add it normally only applies to highly active pages, i.e a small percentage but most likely the pages you spend most of your time on. Regards, SunCreator ^(talk) 00:28, 16 January 2011 (UTC)

I agree, and would add that the "someone else" idea exhibits hostility to editors whose only "flaw" is that they have discovered something. Minasbeede (talk) 03:41, 16 January 2011 (UTC)

There is no problem with an editor citing his own work, as long as that work has been reliably published. The difficulty comes when the editor goes beyond his published work and includes an interpretation, analysis or conclusion that was not explicitly included in his published work. This is common, as the self-citing editor knows background information that was not included in the source. And doing that is OR. Also... the editor who is citing his own work may not be able to stay neutral about the topic. There is the temptation to try to "prove" that his views on the topic are "right". However, that is a WP:NPOV issue and not an OR issue. Blueboar (talk) 16:46, 16 January 2011 (UTC)

Going beyond is not a temptation confined to those citing their own work and, as you rightly point out, that would be an NPOV issue anyway, not an OR issue. So I think we agree: it may be an issue, but it is not an issue that should be dealt with in WP:NOR. Deal with NPOV in NPOV.

Your "explicitly included" makes me uncomfortable: see my wet shoes example above. Scholarly editors (in physical science, anyway) might well remove extremely obvious observations or conclusions from a submitted paper on the specific grounds that the observations or conclusions are obvious and serve primarily to make the paper too long. Removed precisely on the grounds that it is unnecessary to state in a scholarly paper something that is obvious. Wikipedia, according to some, should reject the most very obvious things because they do not appear in any reference, ignoring the fact that the reason something does not appear in a reference is specifically that it is glaringly obvious. (I find it a little petty and a little arrogant to say "Well, go find it in a reference anyway.") An encyclopedia is a vehicle for the dissemination of knowledge. I find it entirely appropriate for an encyclopedia, including Wikipedia, to coalesce existing ideas in a way that better informs the reader. I do not find it jarring for an editor to write in a manner that assumes everyone realizes that directing the stream from a garden hose at ones shoes will make them wet - or will realize so once the idea is presented and the reader has time to think about it. Rather than a reviewing Wikipedia editor asking himself "Is this text OR because it makes a conclusion" I think the editor should instead ask himself "Is this text making an obvious conclusion or a non-obvious one?" If non-obvious, then removal may be appropriate. Even if obvious a conclusion that tends to turn attitudes away from accepted wisdom (which all recognize may be faulty) should not first appear in Wikipedia - that would be OR. (That all recognize accepted wisdom may be faulty does not alter the basic nature of Wikipedia: it is expressly not a venue for new ideas or interpretations to first appear. That is a choice; it is a wise one.)

I believe the article on honeypots, in discussing open-relay honeypots, still indicates that the operator of an open relay honeypot can examine trapped spam and potentially discover useful information. I put that there. The open relay honeypot (when it was a powerful anti-spam technique) did capture voluminous quantities of spam, the operator could look at the spam and see exactly what it contained. (And I did that.) Stringent application of NOR would require, I think, that the fact that the operator could look at the spam and see what was in it must be stated by referring to some published source. Doesn't that border on ludicrous? Does it have to be documented that a person could look at spam and see what it contained in an outside reference? Does it have to be documented that spam must, to work, have to provide some method for the recipient to make a reply, to act in a way that benefits the spammer? In addition, the operator can examine trapped relay test messages (the previous example was about spam: these are another of the things open relay honeypots capture) and see factors about those. Spammers in the past used to employ a simple coding within the message headers to indicate, not quite openly, the IP address of the tested relay. As this coding is not instantly obvious I can agree that indicating the coding used ought to have a reference. This is in spite of the fact that, once the coding method is revealed, anyone can fully understand it. (See my talk page for a little more information.)

(Sorry I'm so wordy.) To me the dominant attitude should be one of determining if something in Wikipedia properly conveys information. Wikipedia requires that the information be verifiable. In most cases this does require the existence of an external source: I agree, I'm content. I assert that there are some things which are obvious, non-novel, and non-controversial which should possibly be allowed without insisting on an external source. ("Possibly" is there to acknowledge that there can be borderline cases and in some of those the proper action is removal.) Not all synthesis nor all reasonable assumption is OR. Where synthesis or assumption smacks of OR - or seems to - whack it. The offense is OR. (But be open to reason in the article talk pages. As it says somewhere, assume good faith on the part of the other editor.) Don't follow the false syllogism: Some OR arises from synthesis, this is synthesis, therefore it is OR. (Which, unfortunately, seems to be thrust into present policy.) Minasbeede (talk) 18:30, 16 January 2011 (UTC)

I'm a published author, I'm a Wikipedia editor, and on rare occasions I have cited myself in an article (with a note on the talk page pointing out that I've done it). I'm against the change that SlimVirgin proposes because

1. it wouldn't solve any COI problem. The real problem arises when editors cite themselves (or their friends) while concealing their involvement. The change in policy would only affect those who are open about their involvement.
2. it would encourage the anti-expert bias that is observable, now and then, on Wikipedia. This bias is a bad thing and it ought to be discouraged. Wikipedia will always need contributions from people who know (including people who have published) and it should continue to welcome such contributions.

I usually write here on topics on which I haven't published, so this change would very rarely affect me, but my observation suggests that the effects would be altogether bad. Andrew Dalby 14:10, 17 January 2011 (UTC)

Synthesis (WP:SYN)

I added the following paragraph to the synthesis section. Was that ok?

“ On the other hand, this principle should be used with caution and not as a way of censoring sourced facts based on a specific POV. Wikipedia is not the place for high-level analysis of results, but do not delete sourced text and rather obvious conclusions without previous discussion, just by referring to this principle. If for example source A says one thing, and from B, C and D it can easily be concluded that A apparently is wrong, a discussion on the talk page is recommended on what is the beast way of dealing with this; if A should be removed, if a low-level synthesis is acceptable or if the information should be organized in another way. ”

Mange01 (talk) 12:18, 7 January 2011 (UTC)

Sounds like there are some very good ideas in there. But it's confusing to read and any significant change to a major policy should be discussed here first. North8000 (talk) 13:37, 7 January 2011 (UTC)

Synthesis on Wikipedia is a massive OR problem. This paragraph does not make the issue clearer. Fifelfoo (talk) 13:41, 7 January 2011 (UTC)

I found the paragraph confusing. I'll remove it if it isn't already gone. Dmcq (talk) 14:02, 7 January 2011 (UTC)

It sounds to me that they're trying to put in something about neutral point of view when there are conflicting opinions in different sources. The WP:NPOV policy talks about that. There is also the problem sometimes that one source is just plain wrong which is a bit of a pain when someone gets a bee about adding it, but fortunately it doesn't happen too often, that requires discussion and common-sense and so basically invoking IAR Dmcq (talk) 14:57, 7 January 2011 (UTC)

Thank you for good comments! The background to my suggestions is that recently I have noticed several times that WP:SYNTH is used as a way of getting rid of sourced facts that someone does not like, for example that contradicts the majority POV among the authors of an article. I.m.o. WP:SYNTH should normally only be used for removing new conclusions, not for removal of sourced relevant facts. If you have another argument for removing sourced facts, then that argument should be presented. ANother point is that praxis shows that this principle is typically only used in controverial issues.

Suggestion for new formulation:

“ On the other hand, this principle should be used with caution to avoid the suspicion of POV based censorship of sourced and relevants facts. Original research such as new high-level analysis and new conclusions may be removed by referring to this principle, but the sourced facts that the conclusions were based upon should not necessarily be deleted. And if sourced facts are deleted, other arguments than WP:SYNTH should be used.

Praxis shows that on non-controversial issues, rather obvious new conclusions is often accepted. [Alternative formulation: If concensus can be reached, low-level synthesis is acceptable, for example non-controversial rather obvious new conclusions and summaries of the sources that do not constitute original research.]

”

Mange01 (talk) 02:20, 10 January 2011 (UTC)

I have long been a supporter of trivial synthesis provided that consensus can be reached, albeit under a different context. In fact, a form of it is already permitted under WP:CALC. In this regard I have previously stated that current guidelines completely disrespect the context of a topic and disregard what may indeed be a triviality facing the complexity of the topic. For example, even though WP:CALC allows trivial calculations like, let's say, a+b-a=b, this may be non-trivial for an article discussing the algebraic axiomatic system.

For example, trivial synthesis may be based on simple logical relations such as transitive and total relations and basic rules of first-order logic. For example, to take an example from a current FAC, if there is a reference that says that something is true for country A, another for country B, a.s.o. till, let's say, G, then it should be permissible to synthesize that something is true for "several countries"^{[refs A-G]}. If all countries are in Europe, it should be permissible to say "several European countries". Such examples can be found in current featured articles, even though current policy precludes this.

In fact, if such simple rules are permitted, it is impossible to conclude that both fact A and not A can be true at the same time. However, given that current policy disregards truth – "the threshold for inclusion is verifiability, not truth" – I don't see how this can be solved. ;) Nageh (talk) 14:38, 10 January 2011 (UTC)

General question: relevant info not mentioning topic

I'm pretty strict about WP:OR and Synthesis myself, but now find three places where a factoid or opinion which does not specifically mention the topic would be useful, though I'm not hell bent on editing either into the relevant article, and don't think I'd get much opposition if I did, so it's not a WP:ORN issue. (However, it may have been an issue in earlier articles, but can't think of example off hand.)

Examples: 1) the fact that military technologists have tried to put cameras/microphones on flies, to show there is some basis in paranoia about "animal spies." 2) a well known sociologist's opinion that the amount of sociological or political power and control can be difficult to guage, in an article about power and control by a specific group in a specific industry. 3) adding a reference where a book title names a person as a celebrity to make sure I can fit that person into a celebrities section of an article, even though book may not touch on the topic for which he is being discussed.

Is editor consensus the bottom line on this sort of thing, or is it just a plain No No?? And does something about "relevant facts/opinion" found in WP:RS not directly related to subject of article may NOT be synthesis need to be mentioned? Or would that be an open invitation to synthesizers, and we have enough problems with them already. Or is this someting that is better discussed in WP:Verifiability? CarolMooreDC (talk) 16:19, 9 January 2011 (UTC)

I think it would run against two principles, NOR and NPOV. In the NOR sense, it is making an assertion or connecting facts that reliable sources discussing the topic do not. In the NPOV sense, our fundamental rule is to represent a summary or depiction of topics as they appear in documents, reputable sources. It causes us to exclude views and information presented by only a tiny minority of sources (as "undue weight"). Observations and ideas not included in any source would certainly cross the line, in that light. --Vassyana (talk) 18:22, 16 January 2011 (UTC)

Thanks for your comment. 1 and 3 have been dealt with. I found a source directly relevant to topic 2 that says "it's hard to gauge power and control on this issue" so maybe just throwing in a footnote saying well known sociologist agrees generally would not be too out of line? (I doubt other editors would call for it to be taken out in any case.) At this point I agree that we shouldn't put in any wording in this article that would mention such a thing, encouraging people to slip down the slippery WP:OR slope. CarolMooreDC (talk) 19:49, 17 January 2011 (UTC)

To give the specifics that Carolmooredc does not, and for obvious reasons does not, the topic she's discussing is Jewish representation in the media. She is on record saying that she believes that the media is "owned and/or controlled" by "mostly Jews," and has been for at least a month ringing the changes, trying to find some Wiki-friendly way to inject some variation on that sentiment into an article about the antisemitic myth that Jews run the press. Read the relevant talk page and you'll see. Opposition has been unanimous at every gambit she has tried so far, and this is just the latest angle that she's trying to play. Spaceclerk (talk) 05:28, 18 January 2011 (UTC)

Do OR rules allow counting or arithmetic?

1. The obituary of Charles Coulson in Biographical Memoirs of Fellows of the Royal Society of London ends with a bibliography of research papers that he wrote. This extends from page 116 to 134 in the pagination of the issue of the Memoirs in which it appears. The papers are numbered [1], [2], ... By scrolling to the final page, I see the final bracketed number is [444]. Is it OR if I state "Coulson was the author, or a co-author, of at least 444 papers in peer reviewed journals" ?

2. Suppose that to prepare some insertions for a WP article, I access the electronic catalogue of holdings of a county library system, and search for books by John Henry Smith, by retrieving records for J.Smith, John Smith, J. H. Smith, John H. Smith, John Henry Smith and J. Henry Smith, cutting and pasting the information about the books that I think are by the John Henry Smith in whom I am interested, and deleting redundancies caused by multiple copies and editions. Then I determine that the number of records is 7, by counting, (or asking a 7-year old grand child to count) and state, in the article, that John Henry Smith published 7 or more books. Have I performed unallowable WP:OR?

3. Suppose I now access Web Of Science or Scopus, and search for papers by John Henry Smith. The system displays the list of citations that it finds. Also it displays "37 items found" (or words to that effect) at the head of the list. I scroll quickly through the titles. All contain distinctive words that are consistent with the work of the John Henry Smith in whom I am interested. I include the statement "John Henry Smith published 37 or more papers in peer reviewed journals." Have I broken WP:OR rules?

4.Now posit the following variation on 3. There are 215 hits, and a quick inspection suggests that 73 are by the John Henry Smith in whom I am interested. I include the statement "John Henry Smith published 73 or more papers in peer reviewed journals." Have I broken WP:OR rules?

5. Revert to case 3. The short information for each hit displays the number of times the paper has been cited. Suppose I add these and get the result 547. Can I include the statement "An indication of the impact of Smith's work is the total number of citations to his published papers. Based on a Web of Science search, this exceeded 540 at the end of the year 2010". Or does the action that enables this statement constitute OR? (And does the mention of citation count as an indicator of scientific impact violate verifiability rules by omitting a reference to article that justifies the action. Then, if I include reference to article that advocates use of citation count as criterion of impact, do I violate WP:NPOV rules. Then, if I just fall back on a string of quotations, do I violate copyright, inside and outside the world of WP?

6. Suppose I download the sets of citing papers for each of Smith's papers, and form the unions of these. This leaves a set of 511 citations. Can I include the statement "An indication of the impact of Smith's work is the total number of papers that cite his published papers. Based on a Web of Science search, this exceeded 500 at the end of the year 2010". Or does the action that enables this statement constitute OR?

7. If a reference source states "Smith enjoyed over 99 birthdays" is it OP to state "Smith lived to over 100"?

8. If a scholarly source states "Umbriago's first reliably dated large scale canvas was completed in 1503, and his last in 1563" can I state "Umbriago was at least 70 when he died", by making the tacit assumption that he was at least 10 when he started to paint large scale canvases. Would this violate WP:OR?

9. Do WP:OR rules permit a statement that locates a building by moving the cursor onto its depiction in a Google Earth session, and transcribing the relevant numerical data displayed at the foot of the screen?

10. Do WP:OR rules permit a statement that gives the distance between two buildings "as the crow flies" based on the use of measuring feature of Google Earth?

11. Do WP:OR rules permit a statement that gives the altitude of an object by transcribing the value shown in a Google Earth session?

12. Do WP:OR rules permit a statement that gives the difference in altitudes of two objects by calculating the difference of the values shown in a Google Earth session, using elementary school arithmetic?

13. Do WP:OR rules permit a statement that gives the inclination of a flight of stairs between two objects by feeding the positions and altitudes shown in a Google Earth session, into a calculation that uses division as well as subtraction?

14. Would WP:OR rules be violated by a statement that gave the inclination transcribed from a resource of Google Earth that provided this directly?

I really want to play by the rules, and there is a learning curve that may be too long for someone my age to climb just by reading the guidelines. Michael P. Barnett (talk) 22:52, 9 January 2011 (UTC)

The best way to play by the rules would be to bring specific examples here when a question arises, since circumstances often help determine what is and is not appropriate. That being said, I'll give you my opinion:

• 1) I think that would be ok, but you might explicitly mention the source in the text of the article in addition to providing a reference, something to the effect "Charles Coulson's bibliography in the Memoirs of Fellows of the Royal Society of London lists him as author or co-author of 444 articles".
• 2 - 6) These, in the general sense, I would regard as OR, since you're doing the research.
• 7) I would simply use what the source uses.
• 8) OR, you can't assume things.
• 9-14), We'd have to determine if Google Earth is a reliable source, and I'm not sure it is. Some of these cases I would regard as OR, such as 12 and 13.

But again, we'd want to work with specific cases as they arise, and what you have above is just my gut feeling in general and thus pretty nigh worthless. --Nuujinn (talk) 23:47, 9 January 2011 (UTC)

I think a number of things you said there would probably be allowed under CALC and commonsense. However they are mainly cruft, if something is interesting somebody will have written about it, they are original reasearch in the sense that nobody ever wrote anything about them. They sound uninteresting to me. The policies are not hard rules, they are what people have written down as what people do in aid of producing the encyclopaedia. Filling wikipedia with cruft does not make it into a good encyclopaedia. You should try and stick to summarizing what people have written rather than try doing research. Dmcq (talk) 23:50, 9 January 2011 (UTC)

In reality, summarization is needed for good writing and information articles, but is technically a violation of WP:NOR. The same for absolute "no brainer" deduction. When that is very obvious and absolutely uncontested, the former is the way that the higher level statements in articles are written. Written by someone who has learned from many many sources, and cited using sources that support the statement but don't explicitly say it. When it's not obvious, and is challenged by someone who sincerely questions the accuracy of the statement, then wp:syn, wp:nor wp:ver provide the quidance to take care of things. When none of the above apply (as when there are real world opponents on the subject), then the system breaks down and the imperfect rules are used for POV wikilawyering towards POV's, and you have the eternal mess that wp articles on contentious topics are in. North8000 (talk) 16:29, 10 January 2011 (UTC)

I agree with Dmcq's analysis. That said, the title of the section is wrong. What you are doing is counting, but you assume that the author is really always the same person. Even if you verify that by comparing subject, publication date, etc., you are doing original research. Simple arithmetic is permitted by WP:CALC.

Regarding Dmcq's statement that "if something is interesting somebody will have written about it"... honestly, I'm reading it so often that it is starting to make me sick. With a little knowledge about combinatorics it is simple to assess that simple application of transitive or total relations amounts to a practically infinite number of very trivial synthesis that most likely nobody will ever have written about. For example, take the deeply structured hierarchies of plant or animal categorizations in biology as an example. Will you be able to find a source that states that Dichelia clarana belongs to the suborder Glossata even though it is known that Dichelia clarana belong to the family Tortricidae and the latter are a sub-category of Glossata? (Now you may certainly find sources that list the whole hierarchy, but I hope you get the point.) See also my comment at #Synthesis (WP:SYN). Nageh (talk) 18:18, 10 January 2011 (UTC)

And hardly any of that is worth sticking in. What I said isn't true as a fact but it is true as far as Wikipedia verifiability is concerned. We as editors can't just go sticking things into Wikipedia without some external proof it is interesting, which is provided by finding a reliable source that writes about it. We're supposed to be writing an encyclopaedia, not doing original research. Dmcq (talk) 18:40, 10 January 2011 (UTC)

I fear I can't follow you. What I'm criticizing is that a strict ban of trivial synthesis is sometimes plain ridiculous, may preclude a good writing style, and for that reason is even ignored in current featured articles! I gave a specific example in section #Synthesis (WP:SYN). I am not criticizing verifiabilityverification per se, as you may easily conclude given that I agreed with your analysis further above. Nageh (talk) 18:48, 10 January 2011 (UTC)

The WP definition of "OR" (essentially anything not sourcable to wp:ver standards) is vastly different than the real world definition which is "Original Research". :-) So in these conversations, we should make clear which term we are using. For example, "daylight came every day last week in Pheonix" is WP:OR (there is probably no source [much less wp:rs]that said that; one would have to synthesize it from sunrise records). but is not "Original Research" per the real world meaning of the term. North8000 (talk) 19:24, 10 January 2011 (UTC)

I agree there should be something allowing the synthesis where one says some or all the countries in Europe when we have evidence for some of the countries or all of the countries. However I think that should be dealt as a case of summarizing where the individual statements have to all be specified and cited later on. There is I feel a general problem where the leads of articles are disconnected from the article because they directly reference citations rather than summarizing the article. However any change along these lines would have to be very carefully considered and discussed at the policy village pump. There are just too many editors around who try exploiting any crack to introduce some of their own theories and they'll go on and on and on arguing that it's alright and what they're saying is just a trivial calculation or not synthesis when it clearly is their own pet bugbear. Dmcq (talk) 20:18, 10 January 2011 (UTC)

(ec) I was just going to post an example but I see that you understand where I am trying to go. I'll post it anyway. Suppose we have sources that say "Country X has to cut their expenditures due to accumulating debts.", where X = Greece, Ireland, Spain, Italy, etc. Why is it so outrageous to conclude that "Several European countries have to cut their expenditures due to accumulating debts." Specifically, why is it banned to state (1) that they are European countries, and (2) that they are several?

Otherwise, I agree with you that "trivial" synthesis is (too) often misused on Wikipedia articles. Nageh (talk) 20:26, 10 January 2011 (UTC)

Days of a week and dates

I am wondering if I can have a reliable source that lists a timeframe for a show and the days of the week it showed, would it be WP:OR to state what those days were? I'm asking because in the past there has been some dispute with this at FLC with List of Popotan episodes in that at least one editor said it would be OR to say that because the air times could have changed. The source I'm using is one that is updated after the fact though. I was eventually able to find a source to satify him, but the site is dead and Wayback Machine is blocking it because the current owner doesn't allow its bots to search it and its likely to remain that way for the forseeable future. Nor can I find a replacement for that info.陣内_Jinnai 23:22, 9 January 2011 (UTC)

Asking again here. Can I remove the source and should it go to FLRC do i need it or can I use the calander without violating OR?陣内_Jinnai 16:04, 13 January 2011 (UTC)

Are you asking whether you can say "Eppisode 12 aired on Friday the 14th of January" if you have a source that said the program aired on the 14th but does not specify the day of the week? (if so, yes, you can... looking up the day of the week that a date fell on qualifies as a routine calculation), or are you asking something else? Blueboar (talk) 21:26, 14 January 2011 (UTC)

The reverse - I have sources saying between dates July 17, 2003 and October 2, 2003 the show aired every Thursday and thus using a calander to say epidoes 3 was on one date, episode 4 on another, etc.陣内_Jinnai 17:35, 16 January 2011 (UTC)

This falls under the common "simple calculation" rule, in my opinion. A list of dates is a long way to list out the same information conveyed by "every Xday between Y date and Z date". Presuming that, as per normal, episode aired in their normal sequence shouldn't require a source. There's not any "interpretation" going on or new information being "created" to cause an original research problem. --Vassyana (talk) 18:27, 16 January 2011 (UTC)

I agree... I think this does fall under the routine calculation, but there is that nagging quibble that we don't know (from the source) if the episodes were aired in the correct order. That could screw up your calculation. My question is whether you really need to be so precise. Why not just say that the show aired every Thursday between July 17 and October 2, 2003, and leave it at that? (if this is for a "list of" article, include the less precise "Every Thursday" statement in the lede, and omit the specific "Air date" from the list itself.) Blueboar (talk) 18:44, 16 January 2011 (UTC)

As its for a chronological list we do need to be precise. Also, for anime tv series at least they are aired in order and its only noted when the exceptions in airing order are done. Unfortunately, no one documents in the sites when a TV series airs normally because its assumed to be that way.

Removing it from the list isn't an option without losing its FL status, which is why I brought it here since the archived site that did mention the exact dates is now being blocked permeable.陣内_Jinnai 19:46, 16 January 2011 (UTC)

Yes, but does it really need to be a chronological list... or can it be organized by episode number instead? Blueboar (talk) 20:36, 16 January 2011 (UTC)

I'd have to bring it up because we already have episode listing and it would be a radical departure from all of the episode FLs I know, especially anime ones. If other episode FLs can make it with just the episode numbers though, it might be possible. As I said though, I think its expected content to have the orginal air date or release date (for non-broadcast episodes).陣内_Jinnai 04:32, 17 January 2011 (UTC)

Where will WP orthodoxy lead?

Writing an article for WP that meets the standards of high quality encyclopedias exposes the Editor, under present WP practice, to erosion of time and stamina that limits further contribution by having to fend off (or subscribe to) hair splitting interpretations (and misinterpretations) of rules based on WP definitions at variance with accepted usage. Articles are rife that flout the spirit and letter of these rules. Is there a danger that present practices can increase the ratio of WP wrangling and flawed articles, to material that is sound? Are there relevant statistics? Extreme interpretations of Copyright, NPOV and NOR rules are in conflict.

It is possible to describe the kinds of things that need to be avoided in a concise manner. Spelling them out in ever increasing detail, however, is subject to difficulties shown by the amount of technical attention they have received in many disciplines for centuries. It would be unfortunate if efforts by WP to reinvent this extensive body of knowledge hampered WP's efforts on other matters.

I see enough reasonable, informed opinion in the discussion of rules and their interpretation to hope the situation can improve. Is it possible to have a WP principle that leaves sound, non-controversial articles (i.e. that contains no statements that anyone challenges as factually wrong, or egregious editorializing or tendentious) in peace, instead of requesting (or making) changes on the grounds of technical violations of some other principle that can be argued interminably. Has this all been said before? Michael P. Barnett (talk) 03:36, 11 January 2011 (UTC)

If an article violates the prohibition on original research, it should be corrected to solve that problem. Just because that article has existed in a stable form for a while doesn't mean it can't be improved. Furthermore, the salient, encyclopedic level points that deserve inclusion in any given article are likely to change over time, as new reliable information is discovered or older information is superseded. If I understand you correctly, you're arguing that someone (who?) should be able to declare articles "in peace," (I assume this means, unchangeable); if that's what you're proposing, I think the vast majority of regular editors would disagree with such a proposition, as continual, incremental change is a pretty fundamental part of how Wikipedia works. Is there some specific example that concerns you that might help illuminate your concern? Qwyrxian (talk) 04:00, 11 January 2011 (UTC)

Absolutely NOT unchangeable. One of my major reasons for contributing to WP is changeability in light of new knowledge, correction of spelling mistakes, rewording or rearrangement for clarity, etc. My perception is that there is always possibility of improvement. Examples of my concern are in ==Do OR rules allow counting or arithmetic?== earlier on this page. The sort of thing I would want left in peace (unless someone found it was wrong) include "John lived to over 100" when quoted source states "John enjoyed more than 99 birthdays". Apart from violating copyright by excessive copying, I want right to express myself this way if, for example, an earlier statement in a paragraph was the direct quote "Tom lived to over 90" and I wanted contrast.

Paradox is this. Finding the statement that "John enjoyed more than 99 birthdays" in a journal article or a book COULD constitute original research, in both accepted usage and OR sense, if it required weeks of searching documents. But who would know? And what if the original source stated the fact in Hungarian? Would I have to leave it in Hungarian? If I found an article about John in Hungarian, I would have to ask someone to translate. How would I know if they translated word by word? By the time people get boxed into spelling out details like "language translation allowed but inference that more than 99 birthdays paraphrases to over 100" we are in a bottomless pit, whether this is done by explicit enumeration or by general rules. Michael P. Barnett (talk) 04:47, 11 January 2011 (UTC) Michael P. Barnett (talk) 04:54, 11 January 2011 (UTC)

Of course, there is the tacit assumption that John was not born on February 29th of a leap year. and is human and not a turtle. Michael P. Barnett (talk) 05:14, 11 January 2011 (UTC)

Let's see:

• "Finding the statement that "John enjoyed more than 99 birthdays" in a journal article or a book COULD constitute original research, in both accepted usage and OR sense, if it required weeks of searching documents." No, I think that's not what we mean by OR. If you find the statement "John enjoyed more than 99 birthdays" in a reliable source, it's fine to use it, regardless of how long it took you to find it. Calling John's relatives, or finding his birth and death certificates would be OR. If you found a source that said John was 50 in 1950, and a second one that said he died in 2000, and combined them to say "John lived to be 100 years old" would also, in my opinion, be OR (SYNTH).
• "And what if the original source stated the fact in Hungarian? Would I have to leave it in Hungarian?" No, if the source is reliable, you'd cite it, english language sources are preferred, but not required. If challenged, you would have an obligation to provide a good translation, and I would suggest that doing so on the article's talk page when adding the citation is good form. And if you cannot read Hungarian, you have a number of options, including Google translate (which should be used with caution). It would be perfectly acceptable to ask another person to provide a translation, and if you include both the original and translation of the source material on the article's talk page, others can check it for accuracy--the key issue here is that the information be verifiable.

Also, if you want to say "John lived to over 100" based on a source which states "John enjoyed more than 99 birthdays", and were challenged on that point, can you point to where that happened? You are casting these as examples, and it's easier to deal with particular cases since we have context for those. For example, I've enjoyed many birthdays which were not my own.... --Nuujinn (talk) 10:39, 11 January 2011 (UTC)

Rigorous expression, example 1

"Wikipedia does not publish original thought" -- should this be "Wikipedia does not publish any original thoughts of the Editor who is making the contribution"? Nobel prizes are awarded for work that must include original thought. Present wording can be (mis)interpreted to exclude, from WP, reasons for awarding a Nobel Prize. Also it begs the question of what is "original" and what is "thought" -- decidedly not trivial. I am not being facetious or hostile or denying the need for rules, in particular addressing the intent of NOR-- just trying to illustrate the difficulty of wording, and where literalism can lead. In a talk that he gave in 1945, the then Mr. Justice Birkett, later Norman Birkett, 1st Baron Birkett, said "the law should be written in such a way that someone reading it in good faith can understand it, and someone reading it in bad faith cannot misunderstand it." Finding an adequate definition for "trivial" synthesis is an enormously difficult, technical problem. It depends on the context -- what is trivial in one context (elementary school algebra) can be very challenging in another (foundations of mathematics). I think we have to fall back on general intent plus common sense and good faith. And defining those is not easy. Michael P. Barnett (talk) 01:37, 13 January 2011 (UTC)

You understand the policy correctly. However, I don't think there is a need to rewrite the sentence you quote. If you put it in context with the rest of the policy, it is clear that is what we mean. Blueboar (talk) 21:30, 14 January 2011 (UTC)