Frequently Asked Questions (FAQ) | |
---|---|
|
![]() Archives |
---|
Threads older than 14 days may be archived by MiszaBot II. |
Contents
- 1 Digital Dream Door deletion
- 2 Academic consensus
- 3 Semi-protected edit request on 1 March 2016
- 4 RfC: Does the section " Vendor and e-commerce sources" still reflect the consensus?
- 5 Archive or availability requirement for text sources
- 6 Protocol question
- 7 Suggestion to rewrite "nothing . . . that is not citable"
- 8 sources which make clear "guilt by association" arguments
- 9 Should the EL guideline and this guideline be reviewed to change to common language where appropriate?
Digital Dream Door deletion
Hello,
I've deleted Digital Dream Door everywhere I could find it because it's not a reliable source. It's an amateur website using subjective and vague methods for building up their lists and gathering their information.88.182.48.88 (talk) 11:03, 27 January 2016 (UTC)
Academic consensus
There are several issues with the academic consensus section (WP:RS/AC). First, it is not clear from the section when WP editors should use general statements that are referenced but not attributed in the article text (e.g., "Most scientists support hypothesis XX") versus identifying particular, named sources. The academic consensus section states "The statement that all or most scientists or scholars hold a certain view requires reliable sourcing that directly says that all or most scientists or scholars hold that view. Otherwise, individual opinions should be identified as those of particular, named sources." It is not clear what "Otherwise" refers to. What is the criteria for using a general statement versus identifying particular, named sources? I think the guideline should make it clear what the decision criteria are for this matter. Second, the academic consensus guideline does not make a distinction between statements made by reliable sources that consist of a single sentence ("Most academics believe that global warming is an unproved theory") and statements made by reliable sources that summarize the results from a large, systematic review ("Smith et al reviewed 5,000 peer reviewed journal articles on global warming and climate change from 1990 to 2015 and found that 95% of the articles supported the global warming hypothesis."). Third, the academic consensus section does not give guidance on how to deal with competing claims about academic consensus. What if Jones states that "most scientists do not support the global warming hypothesis" (which refers to all scientists from all scientific fields, including scientific fields in which scientists do not have expertise in climate change science) and Smith states "most scientists in fields pertaining to climate change support the global warming hypothesis." Fourth, what if there is a divergence between the academic consensus in academic discipline A and academic discipline B? (e.g., "Almost all religious scholars support hypothesis X regarding the Bible. However, most mainstream historians do not support hypothesis X.") Fifth, do we give more or less weight to competing consensus claims depending on the reliability of the author and publisher? (e.g., Smith claims that there is a consensus for hypothesis X on climate change in a book published by a general publisher that publishes a range of commercial trade books versus Jones who claims that there is a consensus for hypothesis Y on climate change in an Oxford University Press book). OnBeyondZebrax • TALK 15:00, 13 February 2016 (UTC)
- Also note that "consensus" (especially in social sciences, but actually for all academic uses of "consensus") often changes rapidly - thus what is "consensus" in 2010 may well be not the consensus in 2020. And, in general, scientists do not "vote" on truth - they simply agree or disagree that a particular theory is more apt to be correct or less apt to be correct in forecasting the results of future empirical observations. One splendid example is the "number of hurricanes forecast in the coming year" which, over a period of two decades, was less accurate than forecasting the exact same value each year over the two decades <g>. For historians, the record is even worse, "revisionism" is rampant in broad swaths of historical commentary. Einstein's proposals concerning energy and gravity is an example where we only now have empirical observations appearing to conform with his position. Collect (talk) 17:00, 13 February 2016 (UTC)
- We report the current consensus regardless of how we think it may or may not change. Returning to OBZ's example of climate change there are multiple sources of extraordinarily high quality that explicitly state what the "consensus" view is. (Granted, not everyone accepts that the pointy-headed intellectuals at the world's national science academies just might be more credible than Ted Cruz.<g>) But not every case is so clear cut, so we have to be very careful. As an aside, the hurricane forcasting example is inapt given that there is no consensus within the relevant scientific community regarding how best to forecast hurricanes. Shock Brigade Harvester Boris (talk) 17:15, 13 February 2016 (UTC)
Semi-protected edit request on 1 March 2016
128.118.134.7 (talk) 16:53, 1 March 2016 (UTC) http://www.pennlive.com/midstate/index.ssf/2011/11/paterno_praised_for_acting_app.html
Reliable source: Pulitzer Prize author who broke the story.
- This page is a guideline about how to identify a reliable source. It is not a list of reliable sources. So the source you mention has no place on this page. Jc3s5h (talk) 17:15, 1 March 2016 (UTC)
RfC: Does the section " Vendor and e-commerce sources" still reflect the consensus?
WP:RS#Vendor and e-commerce sources says: Although the content guidelines for External links prohibits linking to "Individual web pages that primarily exist to sell products or services," inline citations may be allowed to e-commerce pages such as that of a book on a bookseller's page or an album on its streaming-music page, in order to verify such things as titles and running times. Journalistic and academic sources are preferable, however, and e-commerce links should be replaced with non-commercial reliable sources if available.
Bgwhite (talk · contribs), an editor whose work I have seen and respect, has been removing supporting references to Amazon that provide reference support for information such as book titles and publication dates. Examples from my watchlist are:
- Magic Tree House (series) ( | talk | history | protect | delete | links | watch | logs | views); and
- Laura Schwendinger ( | talk | history | protect | delete | links | watch | logs | views).
The references were already present in the Magic Tree House article; I didn't add them. I did add the reference to the Schwendinger article; the material it supports was already there, but unreferenced, and on a quick Google search, I could find no better reference.
I've pointed out this guideline, and that sources should be replaced; but can still be used until replaced.
BGwhite's approach is to simply delete the references, with comments such as "Amazon IS NOT a valid source. Period." I disagree; I think it is acceptable edit behavior to either replace the Amazon cite with a better cite, or add {{better source}} to flag it for another editor, but not to delete the cite and leave it unreferenced.
On his talk page, BGwhite says "It says when a better source is available, use it instead of the commercial site." I'm good with that, but these edits are not substituting a better source; they are not using a better source. They are simply removing the existing source.
I realize for the Magic Tree House article, I could probably google-search and come up with something (and indeed, for the book series, one thing I'd like to do when I have the time is to add an ISBN-based footnote for each book); but I'm thinking of the bigger picture here, and the pattern of deleting sources and leaving material unsourced as a result.
All this basically calls into question this guideline. Should Amazon cites be deleted without replacing with a better site? Should this guideline be updated? TJRC (talk) 19:02, 1 March 2016 (UTC)
- TJRC, I can say from many many encounters with Amazon's data over a period of years when I was working on entering bibliographic data and verifying other people's entries at the ISFDB that Amazon was notoriously inaccurate in reporting exact publication dates. It was usually accurate as to the year, but is simply not to be trusted beyond that. It was also notoriously inaccurate on page counts, often rounding to the nearest power of 2, or simply filling in apparently random figures when a publisher had not supplied a page count. It was also commonly inaccurate on subtitles, presenting series names (often series designations of its own devising, never used by the author or any other source) as sub-titles. Amazon also has a habit of citing current ISBNs on entries about out-of-print editions of books. These issues do not apply to other commercial book-selling sites (such as Barnes and Noble, or most publisher's sites), they are quite specific to Amazon. Therefore I think Amazon should not be considered a reliable source, not because it is commercial, but because it in particular has a demonstrated lack of accuracy. DES (talk) 16:52, 26 March 2016 (UTC)
-
- Thanks, DESiegel. Can you give your input on the questions Should Amazon cites be deleted without replacing with a better site? Should this guideline be updated? TJRC (talk) 19:04, 26 March 2016 (UTC)
- My view, TJRC, is that there is no need to change the guideline on these matters, but we might want to have a discussion, perhaps at WP:RSN, on the reliability of Amazon in particular. This might lead to a consensus to limit the use of Amazon in particular, much like the consensus that exists not to rely on the IMDB except in very limited ways. i think that the best practice would be to replace a cite to Amazon with a cite to another source, even another bookseller. But if no other source can be found, that may be a further indication that the Amazon content is not reliable (when it is reliable, Amazon content can usually be confirmed at other sources) and so should be removed. That, however, is a fact-bound and circumstance-specific decision, to be made on individual article talk pages, or perhaps at AN or ANI if an editor is routinely or massively removing Amazon cites and not replacing them with easily available (i.e googleable ) alternate cites, after discussion. That is my view. DES (talk) 19:24, 26 March 2016 (UTC)
- Thanks, DESiegel. Can you give your input on the questions Should Amazon cites be deleted without replacing with a better site? Should this guideline be updated? TJRC (talk) 19:04, 26 March 2016 (UTC)
Archive or availability requirement for text sources
In the discussion at Wikipedia talk:Citing sources# Citing a blast email three has been disagreement as to how available a text source must be for it to be considered reliably published. WP:RS#Definition of published says that:
However, audio, video, and multimedia materials that have been recorded then broadcast, distributed, or archived by a reputable party may also meet the necessary criteria to be considered reliable sources. Like text sources, media sources must be produced by a reliable third party and be properly cited. Additionally, an archived copy of the media must exist.
Moreover Wikipedia:Published says:
All reliable sources are, by definition, both published and accessible to at least some people. Sources that are not published (e.g., something someone said to you personally) or not accessible (e.g., the only remaining copy of the book is locked in a vault, with no one allowed to read it) are never acceptable as sources on Wikipedia.
However, Wikipedia:Published is an essay, not a guideline or policy.
In the WT:CITE thread, some editors asserted that an out-of-print book not held by any library or offered for sale by any used bookstore could nonetheless be a reliable source if a single editor asserted possession of a copy, even if no other editor was in a position to verify this. I disagree.
I would like to add wording to WP:RS, perhaps in WP:RS#Definition of published to make a requirement equivalent to the "archive" requirement for broadcast media, to say that a source is not considered published unless at least one library holds a copy, or it is available for sale, or there is some other method whereby a reader could, even if with some effort or expense, obtain a copy to verify the citation. DES (talk) 17:07, 26 March 2016 (UTC)
- Four comments. (1) The idea that a book must be held by a library or an archive or currently available somewhere for purchase goes against the common sense notion of the word "published". So I would suggest rather than trying to re-define the word "published" in a counter-intuitive way it would be better to just add as a requirement for what counts as a "reliable source" this extra condition. So, for example, it would be better to change the first sentence of the "Overview" section to say "Articles should be based on reliable, third-party, published, currently available sources with a reputation for fact-checking and accuracy." Then there could be a little section on what counts as "currently available" which specifies the one library, one archive, one copy for sale idea.
- (2) Changing the wording on this page is fine, but this is just a guideline. Unless there is also a change on the page WP:SOURCE, which is a policy page, making a change here won't really settle the matter. The policy page still will not say anything about the current availability of text being a requirement for counting as a reliable source, and so since policy pages trump guidelines making a change here won't change policy. I would recommend seeking a change in the policy page for this change to have real force.
- (3) The suggested change is unlikely to make a real practical difference. We have all, I am sure, encountered editors who are quite insistent about edits they are sure are correct, so imagine just how more determined an editor would be if they actually are holding in their hands a copy of a book that was published that supports the information they want to add to an article. If other editors object to adding the information and explain the "one library, one archive, one copy for sale" condition it would be quite simple for the owner of that copy of the book to put it up for sale online with an asking price of $1m, thus satisfying the "currently available" criterion. A condition that easy to satisfy is really no condition at all.
- (4) It is important not to confuse worries that an editor is mistaken about a source or the reliability of a source with worries that an editor is just making stuff up (and that no such book even does exist). In the former case, because text can be quoted verbatim and information about the publication history of the book can settle whether the book is likely to have the facts right and whether the editor has made a mistake in reading the source, requiring copies be available for others to check is not really that important. Honest editors of good faith don't need any change in guidelines or policy. However in a case where some editors believe that another editor is fabricating a source that does not even exist or is attributing words to a source that it does not even contain them the issue isn't reliability of the source at all. Requiring an editor to prove that a book exists is different from requiring the editor to use a source that is reliable. Insofar as there might be concerns about falsification of sources, changing the criteria for reliability is not the way to go about it. 99.192.48.16 (talk) 17:43, 26 March 2016 (UTC)
-
- The issue here isn't reliability... It's Verifiability... Any document is reliable for a quote or close descriptive statement as to the content of the document ... But if a reader (or someone else, acting on the reader's behalf) can not gain access to the document, then the document is not Verifiable. and Verifiability is a policy. Blueboar (talk) 18:03, 26 March 2016 (UTC)
- (edit conflict)Hello again, 99.192. You are quite probably right that if such a change is made it should be to impose a "currently available" condition on reliability, rather than making it part of the definition of "published". It should still have the effect that I desire. In regard to your point 4, I am not so much concerned with the editor who completely fabricates a source as with the editor who quotes a source out of context or otherwise misrepresents a source, quite probably doing so in good faith. Without the ability to inspect not only a specific quite but the surrounding text, it is often impossible to see if a source really does support the statements for which it is cited. Editors, particularly unconsciously biased editors, all to often cherry-pick source quotes, and it may take a fuller inspection of the source to see that the quote does not actually support what it is cited for. Also, inspection of a source, particularly a historical source, may well reveal facts about it that cast doubt on its reliability. (For example it may make frequent errors of fact in other matters, it may cite as reliable sources known to be unreliable, it may display clear bias, etc) If it is an obscure source, these may not be evident without such extended perusal of the source. In short, we need the ability for some other editor to verify the source for it to be considered truly reliable. DES (talk) 18:15, 26 March 2016 (UTC)
- The issue here isn't reliability... It's Verifiability... Any document is reliable for a quote or close descriptive statement as to the content of the document ... But if a reader (or someone else, acting on the reader's behalf) can not gain access to the document, then the document is not Verifiable. and Verifiability is a policy. Blueboar (talk) 18:03, 26 March 2016 (UTC)
-
-
-
- DES, all you say makes sense, but how is a used bookstore down the street from me having the only copy in the world available for sale going to help editors check for context? Or, as in my example, an editor offering to sell the book for a price no other editor would want to pay? Satisfying the availability criterion is no way to ensure the source can be read by others. But also, if we are assuming that the editor who has the only copy of the book is editing in good faith, there are lots of way to make it possible for others to check. The editor who owns the book can scan a few pages and send them to others to read. It's probably a lot easier to do that than to travel to New York City to see the only copy of a book located there in a public archive. So I don't see the proposal ultimately being much help. 99.192.48.16 (talk) 18:31, 26 March 2016 (UTC)
-
-
-
-
- (edit conflict)Two things, Blueboar. (1) The policy I linked to (WP:SOURCE) is a subsection of Wikipedia:Verifiability, and that policy is the one that does not include an availability requirement, so you are wrong about what "verifiability" means. (2) You are misusing the term "verifiability". For Wikipedia "verifiability" is a description of claims made in an article, not of the sources used to support claims. So if an article says that John Smith was born with six fingers, that claim is verifiable if there is a reliable source that says he was born with six fingers. To say that a source is not verifiable does not make sense with how the policy uses the words "verifiabile" and "verifiability. 99.192.48.16 (talk) 18:25, 26 March 2016 (UTC)
-
Protocol question
Hi. I am coordinating the latest database for The Wikipedia Library (Alexander Street Press). Where is the best place to announce this latest resource so editors in need can benefit from using it to acquire references for articles they are improving or proposing? Ping me back. Cheers! {{u|Checkingfax}} {Talk}
22:09, 27 March 2016 (UTC)
Suggestion to rewrite "nothing . . . that is not citable"
May I change a double negative to a more direct statement?
- Change from: Because Wikipedia forbids original research, there is nothing reliable in it that is not citable with something else.
- Change to: Because Wikipedia forbids original research, everything reliable in Wikipedia can be cited from another source.
I will change the sentence if two editors say yes (and no one says no). Scenography (talk) 18:10, 2 April 2016 (UTC)
- I would expand this to " Because Wikipedia forbids original research, everything reliable in Wikipedia can be cited from another source, and it is better to cite such a source than to cite Wikipedia directly". DES (talk) 23:11, 6 April 2016 (UTC)
sources which make clear "guilt by association" arguments
On Talk:Vladimir_Putin a claim is made that if a "reliable source" presents a "guilt by association" argument, that Wikipedia is allowed to stated that argument as fact.
The post:
- But we ARE NOT "iterating claims of guilt by association as fact". We are simply reporting what reliable sources say. That's it. If a source says "papers leave a trail leading back to Putin" then that's what we write. What we cannot write is "Putin stashed away millions" - and it would be only in that case that we would violate BLP. But that's not what we are doing here.
The edit made by that poster:
- In April 2016, 11 million documents belonging to a Panamanian law firm Mossack Fonseca were leaked to the German newspaper Süddeutsche Zeitung and the Washington-based International Consortium of Investigative Journalists. According to the Guardian, the name of Vladimir Putin "does not appear in any of the records", but it and other Western media published lengthy reports about three of Putin's alleged friends on the list.[356] According to the Panama Papers leak, close trustees of Putin own offshore companies worth two billion US-Dollar in total.[357] The newspaper Süddeutsche Zeitung regards the possibility of Putin's family profiting from this money as plausible.[358][359] According to the paper, the US$2 billion had been "secretly shuffled through banks and shadow companies linked to Putin's associates", and Bank Rossiya, previously identified by the U.S. State Department as being treated by Putin as his personal bank account, had been central in facilitating this.[360][361] A significant proportion of the money trail leads to Putin's best friend Sergei Roldugin. Although a minor musician, and in his own words not a businessman, it appears he has accumulated assets valued at $100m, and possibly more. It is been presumed he was picked for the role because of his low profile.[362]
Where the source cited specifies that Putin's name was not in any of the 11 million documents, cited people as being "alleged friends", profiting as "plausible", and the classic comment about a person about whom no charges exist "It is been presumed he was picked for the role because of his low profile". Is this use in a BLP about Putin as a living person an example of guilt by association, and use of rumours? I comment disinterested editors to examine the sources and claims made for such sources, but that is not the question I raise here.
Should WP:RS state openly that
-
- Claims of "guilt by association" are deemed to be "matters of opinion" and not allowed to be stated as fact in Wikipedia's voice
Or is this going to far, or is it insufficient to prevent misuse of opinion sources making allegations of criminal acts? Collect (talk) 15:39, 6 April 2016 (UTC)
Should the EL guideline and this guideline be reviewed to change to common language where appropriate?
I was recently working on improving some references—mainly cases where an editor included a bare link, and I was converting it to a proper reference. In one case I noticed that the person linked a Google search as opposed to a specific entry. That didn’t seem right but I didn’t see any prohibition in this guideline (did I miss it?).
While working on something else, I happen to review the external links guideline, specifically, links to be avoided. There, in item 8, it specifically suggests avoiding “links to individual website searches, search engines, search aggregators, or RSS feeds”. While this is in the external links guideline, I think it equally applies to references.
This leads to two questions, one specific, one general:
- Specific: should this guideline be amended to suggest that references to individual website searches should not be generally viewed as an acceptable reference?
- General: should this entire list, or possibly both guidelines be reviewed so that the two guidelines are coherent—in other words, while there may be some cases where advice applies to an external link or to a reference, but not both, in many cases the advice will be identical and we would be better served if we use the exact same language in both cases. As an additional example I note item 17 suggest avoiding referral links. This advice would also seem to apply to references but I do not see it in that guideline.--S Philbrick(Talk) 15:44, 17 April 2016 (UTC)