Wikipedia talk:No original research

wp:quote

There is a proposal to promote this.174.3.113.245 (talk) 06:26, 29 March 2010 (UTC)

Original Research On Philip Larkin

almost-instinct says on Wikipedia_talk:Quotations#Unable_To_Post: "For the biogs sections I chose quotes that had some relevence to that section of Larkin's life. The other quotes are from popular poems and can stand alone."

We already have an article listing of Philip Larkin's Poems.

His choice of quotes is original research. Can we get more feedback?96.52.92.106 (talk) 03:44, 5 April 2010 (UTC)

Duplication of Verifiability

Much of the lead and sources section are a repetition of the verifiability policy. This duplication confuses and dilutes the principal of No original research.

I suggest that most of the duplication be moved to the Related policies - Verifiability section or removed from this policy.--SaskatchewanSenator (talk) 08:30, 7 April 2010 (UTC)

WP:SYN allows strawman fallacy & false premises

Many people have been troubled by the policy of WP:SYN, to reject synthesis of reliable sources to promote an unpublished, so-called "conclusion". After years of consideration, I have finally pinpointed the problem: the policy of WP:Synthesis allows people to assume untrue conclusions which can form the basis of 2 major logical fallacies:

1. a strawman fallacy, by claiming an untrue conclusion similar to the most typical true conclusion; or
2. an argument from false premises, by claiming any reasonable, but false conclusion, and then proving anything from that false premise (such as "delete all that text").

In particular, when people claim the sources had stated A & B leading to an "implied conclusion" C, that actually empowers the assumption that C is the intended conclusion. Anyone can claim the nature of C, because it need not be stated (big mistake), and thus, C is wide-open to interpretation, as being the assumed, unspoken conclusion. Hence, by allowing any potential assumption, that allowance has opened the door to claiming a false premise, as a false basic assumption to argue against the article's text. By using the logical fallacy of argument from false premises, then anything can be proven; therefore WP:SYN could be legitimately used to prove that almost any text should be removed from an article. For example, consider the following extreme, but legitimate case:

• Fact A: The lake's water level was 75 feet deep last year.
• Fact B: The lake's water level is 50 feet this year.
• Conclusion: "OMG, delete that text because you're saying the lake will be empty in 2 years and they're all going to die!!! Aaargggg!"

Unfortunately, that implied conclusion cannot be rejected (because any unspoken conclusion is allowed by WP:SYN), and hence, quite possibly, arguing from the false premise of an empty lake in 2 years, then perhaps many people would die. Result: per WP:SYN, those 2 facts must be deleted from the article, as the complaint is indeed valid, per policy, for removing that text. The policy has failed because of a critical major mistake: anyone can assume almost any conclusion. Example:

• Fact A: The glass was totally full yesterday.
• Fact B: The glass is half-empty today.
• Conclusion: "Oh no, you're saying the glass will be totally empty tomorrow, oh POV...delete...POV POV...delete delete.

A similar example:

• Fact A: The glass was full yesterday.
• Fact B: The glass is half-full today.
• Conclusion: "Oh no, you're saying the glass will be full again tomorrow, oh no, a full glass with no space for more, oh no, oh no, that's WP:SYN...delete delete delete."

Once again, despite how twisted or rabid the conclusion, those are indeed valid reasons to completely remove the related text.

A related, but more subtle, fallacy would be to substitute a "similar, unspoken conclusion" to be refuted, by condemning that nearby-conclusion using a strawman fallacy. Simply put: a Wikipedia policy must not reject anything based on a user's own assumptions. All policies must deal with what is actually written in an article ("put it in writing"), and not prosecute a "pre-crime" action as if it were based on hard evidence. Absolutely nothing based on assumption should be the basis to reject text: once unbounded assumptions are allowed, then anyone can validly do anything to an article. People often use WP:SYN to remove questionable text. However, because of allowing false premises (and subtle straw man fallacies), the failed WP:SYN policy can also be used by any desperate or frantic person to slant or censor any article, and thus it has been.

Recommendation: Issue a major retraction of WP:SYN, and issue a meta-policy that prohibits any future rejection of text based on a user's mere assumptions, rather than tangible hard evidence, to be compared against policy standards. -Wikid77 01:15, 11 April 2010 (UTC)

Can you point to one real example in a real article where material kept out by WP:SYN as currently written, should be allowed in your view? Crum375 (talk) 01:22, 11 April 2010 (UTC)

• In general, see them all: Special:WhatLinksHere/Wikipedia:SYN, select a page and search for "WP:SYN" or "synthe". No removals are really permanent, and people could later re-add material, so the phrase "kept out" is not accurate, but rather, the focus is that someone wanted information excluded, claiming WP:SYN, during a debate. I fear listing any particular articles here, because many people still think WP:SYN is a valid reason to axe text, so I don't want to list any titles which could be viewed as "marked" for death by WP:SYN fallacious thinking. Some broad examples would be:

• Claiming some particular "Category:" names were too original, as a form of WP:SYN.
• Rejecting a list of related atrocities in an article as WP:SYN which advances some "outrageous" conclusion about societies, even though those events are all actually examples providing some real-world details, as pertinent to the subject (An example could include viewing the "local natives" as "massacred" but all the many white people were merely "killed" by natives trying to defend themselves or their land)
• Demanding that rebuttal of evidence in a trial all flow from one source, lest multiple sources be used to "prove he didn't do it" when a source doesn't explicitly state that conclusion, but rather that the evidence was improperly handled. Gripe: "You're implying he's innocent, when that source just stated 1 item of evidence seemed tainted" or similar wiki-weaseling to prevent the listing of more details.
• Removing tangent statements from articles, based on the notion of how details from other sources were "advancing a cause" rather than just expanding the broader coverage of a subject.

In general, the powers of exclusion, stemming from WP:SYN, are a "POV-pusher's dream" to systematically remove text from an article where all sources don't contain the alleged "conclusion being implied" when listing all the related facts in one section of text. Again, the unfortunate truth is that WP:SYN really does empower censorship of details coming from multiple articles which don't "explicitly state" a conclusion claimed as the nefarious position being advocated. In fact, I suspect that WP:SYN, ironically, is itself an original-research notion of a content policy: what 10 reputable news-broadcasting groups have policies similar to WP:SYN? I doubt there are many, as WP:SYN is just too peculiar. -Wikid77 (talk) 11:48, 11 April 2010 (UTC)

As I suggested above, please provide one specific example, not a group or dummy cases. If you focus on one specific article and one specific point there, and show how the current WP:SYN policy excludes specific material which in your view should be included, we can address it and see if we agree. Crum375 (talk) 13:07, 11 April 2010 (UTC)

I agree with Crum. We do need specifics. It is not difficult to construct hypothetical situations where a policy would be misapplied (we can probably do this with all our policies). The fact that it is possible to misapply WP:SYN does not negate it, and is overshadowed by the more numerous situations where it can and has been applied properly. Blueboar (talk) 13:43, 11 April 2010 (UTC)

Well, taking this on philosophically, I think there is a point here (though I don't currently agree with the conclusion wikid offers - and yes, an example would be useful). The problem is that wikipedia expects, against all common sense and observed evidence, that all editors can and will operate using functional scholarly reasoning, but wikipedia provides no structure for distinguishing between well-reasoned arguments and crapulently-reasoned arguments according to scholarly principles. It's like setting up a collaborative software design team where half of the collaborators only partly understand the computational syntax of the programming language (and a small percentage wonder why all those curly-braces are there), but everyone gets rewarded according to how much they produce. You end up with people being blindly insistent about conclusions drawn from bad scholarship, and no real mechanism to point out that their scholarship is bad, thus reducing the debate to matter of personal relativism (something no scholar would ever in a million years accept). I see highly experienced editors do this all the time, in various forms (usually on fringe articles, where for some reason they tend to fight irrationality with bigger irrationality).

I don't have a solution; I'm just saying that the problem has merits for consideration. --Ludwigs2 14:43, 11 April 2010 (UTC)

When rules are written such that normal practices widely "violate" them you create a potential for abuse. For example, by people who want to make articles POV, or to use articles as a boxing ring, or people with other anti-social tendencies, such as those whose psychological needs require attacking other people and their work rather than creating anything. As Wikipedia matures, the 30,000' view is that overall things are taking a turn for the worse in this area. With out-of-context pieces of WP:OR/Synth being one of the most widely violated rules, it would be a good one to look at with respect to this. North8000 (talk) 15:11, 11 April 2010 (UTC)

• If there is some crucial value provided by WP:SYN, then let's consider alternatives. What if the indicated "conclusion C" were required to be stated? It would only be WP:SYN if a writer added, "The U.N. is thus a total failure." By requiring actual, tangible text which states the disputed conclusion, then a WP:SYN problem could be pinpointed to the existing text, rather than to an "assumed conclusion" which gets used in an argument from false premises. When I hear the "160 wars" after the UN was formed, I might conclude, "I wonder how many wars there were before the UN?" rather than, "Aha! I knew the UN was a terrible, utter waste of money: let the guys with the biggest nuclear blasts decide who rules the remaining fall-out shelters, and who cooks the bodies for food!" or whatever rabid conclusion people worry might be claimed. We need to see what POV problems are arising, and find policies to address those issues directly, not open a false-premises Pandora's box, of diseased thinking, to reject facts from articles which do not "directly state" some assumed nefarious conclusion. WP:SYN should not be used as a crutch to cover for a lack of other policies. WP:SYN enables people to claim some sourced facts are not allowed, because they bolster some "evil" conclusion. Instead, simply require NPOV balance, by the addition of alternate viewpoints. That is the fatal problem with WP:SYN: allowing censorship of facts based on assumptions of misuse. That dog won't hunt to collect real knowledge. -Wikid77 19:26, 11 April 2010 (UTC)

Wikid77, I asked you twice above for one real example of a real article where the current WP:SYN policy excludes material which in your view should be in the article, and have yet to see it. All the hand waving in the world can't replace one simple example showing how the policy fails in a real article. Unless you can actually show where it's broke, there is no point fixing it. Crum375 (talk) 22:16, 11 April 2010 (UTC)

• Examples: The issue should be decided on multiple examples, because focusing on 1 example might imply, in itself, a strawman decision, that the application of WP:SYN could somehow be decided by the outcome of debate over a single article. With that stated, a recent, single example was discussed in "Talk:Submarine#Women" when noting: (Fact A) some Navy officials wanted to have women in submarine crews, while (Fact B) those officials were not qualified in subs and had never served in subs. The claimed "assumed C" was that people with no credentials in submarine operations were not qualified to decide that role of women, rather than just being a statement of fact. Similarly, someone indicated it would be improper to state those officials were all men (none of them women) as another case of WP:SYN. The censorship then became: can't mention those officials had no submarine training; and can't mention they are all men. Meanwhile, no one ever said, "Only women with accredited knowledge of submarines should make decisions about women submariners". The problem of WP:SYN is not specifically about submarine service, but rather the censorship of sourced details: censored because those details might bolster an unsourced, assumed conclusion which no one ever advocated in writing. That violates WP:NOTCENSORED. I suggest to reword and only apply WP:SYN when someone states a specific conclusion, such as: "Only women should decide women's duties" or such. Remember to consider other examples in the 6,590 pages listed (during April 2010) with Special:WhatLinksHere/Wikipedia:SYN. -Wikid77 06:12, 12 April 2010 (UTC)

The implication of mentioning those characteristics of the officials in this context is pretty clear. Unless a reliable source has made that connection, it is original and should continue to be prohibited by this policy. I don't see a strawman fallacy or false premise. Can you choose another example?--SaskatchewanSenator (talk) 08:48, 12 April 2010 (UTC)

I don't think the women submariner example shows any deficiency in the current SYN policy; if anything, it shows the policy working exactly as intended. If the proponents of women submariners share certain characteristics, it is not for Wikipedians to decide what those are and write about them, as that would constitute original research, and more specifically WP:SYN. If all the proponents do in fact share a characteristic, and it is notable and relevant, some news organization or another reliable source would pick it up and publish it, and we could then cite it. In this specific example, the lack of such sources could mean that there are other proponents who don't share the characteristic (for example who spoke to reporters off the record), or other reasons the news organizations don't feel the information is valid or reliable. The whole point of Wikipedia's WP:V and WP:NOR sourcing policies is to only add information which has been published elsewhere by verifiable reliable sources, not to make up, synthesize or imply new material or novel interpretations to suit our POV. Crum375 (talk) 11:49, 12 April 2010 (UTC)

I agree, the example is clearly a case of WP:SYN (the implied conclusion clearly being that those who were not qualified in subs and had never served in subs are not qualified to comment on the subject of women in submarines). Even if there are multiple possible interpretations, what is obvious is that the reader is being led to form a conclusion... that is improper. Blueboar (talk) 12:36, 12 April 2010 (UTC)

Lead

We had agreement last month to reword this a little, but it didn't get done at the time, so I've just added it. It says the same thing, but the writing's clearer. Reproducing below part of the discussion from last month. SlimVirgin ^{talk contribs} 02:43, 12 April 2010 (UTC)

Previous

Current

Wikipedia does not publish original research or original thought. This includes unpublished facts, arguments, speculation, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position. All material added to articles on Wikipedia must be attributable to a reliable published source, even if not actually attributed in the text. This means that Wikipedia is not the place to publish your own opinions, experiences, arguments, or conclusions. Citing sources and avoiding original research are inextricably linked. To demonstrate that you are not presenting original research, you must be able to cite reliable sources that are directly related to the topic of the article, and that directly support the information as it is presented. The sourcing policy, Verifiability, says that citations must be added for any material challenged or likely to be challenged, and for all quotations. "No original research" is one of three core content policies, along with Neutral point of view and Verifiability. Jointly, these policies determine the type and quality of material that is acceptable in articles. They should not be interpreted in isolation from one another, and editors should therefore familiarize themselves with all three.

Wikipedia does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources. It also refers to any analysis or synthesis by Wikipedians of published material, where the analysis or synthesis advances a position not advanced by any of the sources. What this means is that all material added to Wikipedia articles must be attributable to a reliable published source, even if not actually attributed. The sourcing policy, Verifiability, says a source must be provided for all quotations, and for anything challenged or likely to be challenged—but a source must exist even for material that is never challenged. "Paris is the capital of France" needs no source because no one is likely to object to it, but we know that sources for that sentence exist. If no source exists for something you want to add to Wikipedia, it is what we call original research. To demonstrate that you are not adding original research, you must be able to cite reliable published sources that are directly related to the topic of the article, and that directly support the material as presented. "No original research" is one of three core content policies, along with Neutral point of view and Verifiability, that jointly determine the type and quality of material that is acceptable in articles. Because these policies work in harmony, they should not be interpreted in isolation from one another, and editors should try to familiarize themselves with all three.

SlimVirgin ^{TALK contribs} 02:05, 13 March 2010 (UTC)

That looks very good to me. The first paragraph is wonderful. My only suggestion is to replace "but we know that sources for that" in the second paragraph with "but also because we know that sources for that" or "and also because ...". — Carl (CBM · talk) 04:00, 13 March 2010 (UTC)

Very good re-write. Jayjg ^(talk) 02:19, 14 March 2010 (UTC)

That would turn it into a slightly different point.

My suggestion is: "'Paris is the capital of France' needs no source because no one is likely to object to it, but we know that sources for that sentence exist."

Your suggestion is: "'Paris is the capital of France' needs no source because no one is likely to object to it, but also because we know that sources for that sentence exist."

The first says it's okay without a source, and it doesn't violate NOR because we know that sources exist for it. The second says it's okay without a source because we know that sources exist for it. But we often know that sources exist for something and yet we still need to see them. The point here is that it's okay without a source for a different reason, namely that no one is likely to object. SlimVirgin ^{TALK contribs} 02:21, 14 March 2010 (UTC)

The point of the NOR policy is that what we call "original research" comes down to claims that are not already published. So claims that really are already published elsewhere are not "original research" for us regardless whether sources are explicitly cited. Indeed, according to the second paragraph of the policy as it stand, what matters is "you must be able to cite reliable sources" (my bold). We might say that we don't know whether some particular claim has been published before, and so we need to see sources to tell whether the claim is original research. But if we already know that the claim is published elsewhere, then we know we are able to cite sources for it, so it is not original research here. So one reason we know that the claim about Paris is not original research is that we know we can cite sources for it. — Carl (CBM · talk) 02:34, 14 March 2010 (UTC)

Yes, my only point is that the reason "Paris is the capital of France" doesn't need a source is not that we know sources exist for it. We know sources exist for lots of things that we insist on sources for. That we know there's a source may be a necessary condition of not asking for one, but it's not a sufficient condition. The reason we don't ask for a source (per V) is that we know no one will reasonably object to the sentence. The first version of the sentence I proposed avoids all these issues. SlimVirgin ^{TALK contribs} 02:52, 14 March 2010 (UTC)

I don't think that's true. Editors don't usually require citations for statements of plain facts for which everyone knows that sources exist.

I don't think that you haven't taken the logic back far enough. WP:V does not require a source for "Paris is the capital of France" because no one will "reasonably object" to it. Now: Why will no one "reasonably object" to this statement? Well, because everyone knows is sourceable. WhatamIdoing (talk) 19:46, 12 April 2010 (UTC)

• Refinements of wording: I recommend the following slight changes to the wording, to provide clarification of meaning:

• Old: "what we call original research"
• New: "what we call original research" [with italics]
• Old: "doesn't need a source"     [confusing as just "source"]
• New: "doesn't need a cited source"
• Old: "in harmony"   [sounds too easy or too smooth]
• New: "in conjunction"

Should other phrases be adjusted? -Wikid77 (talk) 06:54, 12 April 2010 (UTC)

I suppose the lead is only a part of it, but I think that about half of all of mis-uses of this policy and it's biggest fundamental problem could be solved by merely giving more weight to the "challenged or likely to be challenged" part of this policy (quoted from wp:V). Including adding something on the order of "if the accuracy of a statement is challenged, if it is not directly supported per wp:verifiability, it is to be removed. This adds the hint that a challenge (however brief and unsupported the challenge is) of the accuracy of the statement is required (not merely a claim of "OR" or "Synth") in order to have it removed. Again, the disparity between reality (about 1/2 of the sentences in WP are from knowledge/synth and undisputed rather than from statements in references) and the rule leaves it open to abuse, and such has been widespread. I believe that this subtle change would help fix that.

A better example than "Paris is the capital of France" would be "most people believe that the sun will rise tomorrow". We have no reason to believe that such a poll exists to reference. If someone challenges the accuracy/correctness of the statement, they can remove it for being unsupported. However, there would be nothing implying they can remove it for merely claiming it is OR/Synth.

One could claim that this is the purview of wp:V, not wp:OR, but in reality the subject of wp:or is a subset of wp:v and so it has to deal with that. If it is going to make up what are essentially wp:V rules, it has to do so in a way that will prevent them from being widely abused. North8000 (talk) 11:48, 12 April 2010 (UTC)

I think that there is an advantage to retaining "original thought". We don't want to get trapped in "but it's just a new idea I had, not research!" WhatamIdoing (talk) 19:52, 12 April 2010 (UTC)

The problem I see is that "advances a position" is not well enough defined. I agree with the thought. The problem is that there can be synthesis that doesn't "advance a position" but does clarify or illuminate. The bad effect is that while the wording forbids synthesis that "advances a position" (a good prohibition) the practice too often is the prohibition of all synthesis, even when that synthesis doesn't "advance a position," it clarifies or illuminates. Such clarification or illumination is one of the best attributes of an encyclopedia and ought to be encouraged.

Yes, that can be difficult - but that difficulty inherent in the nature of a (good) encyclopedia. It would be grievously wrong to throw out one of the best attributes of an encyclopedia through the over-application of a rule. The question is "Is this instance of synthesis one which advances a new position or is it one that clarifies or enhances?" That is, does the synthesis illuminate some aspect of a topic or is it an attempt to manipulate sources in order to create false favoritism for an idea? "False favoritism" implies that sources are being misused in order to create flawed support for a "novel" idea.

Perhaps what is needed is examples of valid synthesis, synthesis that does illuminate, does not "advance new ideas." (Here I am strongly attached to valid syllogism: if both the major premise and the minor premise are valid in an encyclopedic sense then I favor the embracing of the conclusion, whether or not such conclusion has been explicitly published elsewhere. This statement does hinge on the meaning of "valid." I embrace meaningful and thoughtful enforcement of "valid" in this context. I am not advocating any sort of loophole that allows actual "original research" to intrude. I do not believe and I think the responsible persons who contribute to Wikipedia similarly do not believe that original research is bad or invalid. The assertion and the policy is, simply, that Wikipedia is not the place for such to appear. Wikipedia and its supporters and maintainers do not assume the burden of vetting original research. All such belongs elsewhere. This prohibition, however, should not be interpreted to forbid all thought. "Thought" and "encyclopedia" are not disjoint. An encyclopedia that excludes thought is a weak encyclopedia.) Minasbeede (talk) 11:24, 13 April 2010 (UTC)