Wikipedia talk:Reliable sources

The origine of title Bhaduri

So far i have have knowledge of my study there was Gandagram called Bhaturia under Parjna area of Bengal where an famous Acharya called Udayan Acharya lived. The name Udayan was so famous that often he called as Udayan Bhaturia and the so after long period of the the term Bhaturia is known as Bhaduri. There are many example of Bhaduri like Raja ganesh of medieval Bengal. There are also example of Bhaduri at Saptadurgaat Chalan Beelat Rajsshahi, Now at Bangladesh. — Preceding unsigned comment added by 14.96.51.236 (talk) 14:51, 29 March 2014 (UTC)

Textbooks...

• 1. "Articles should rely on secondary sources whenever possible. For example, a review article, monograph, or textbook is better than a primary research paper."

Textbooks here should be deleted, since they are almost always tertiary sources. So I move to delete this mention of textbooks from this sentence.

Additionally I would like to open up a more general discussion on textbooks. Textbooks are very commonly unreliable. Certain exceptions exist, for instance "widely recognised standard textbooks written by experts in a field" are reliable as the medical claims section states. However, the textbooks students receive up to highschool, textbooks from general education courses in university, and many textbooks beyond that very rarely fit that description, and if possible preference should be given to a reliable secondary source. Caution should be exercised using most textbooks, and in particular these lines worry me:

• 2. "When available, academic and peer-reviewed publications, scholarly monographs, and textbooks are usually the most reliable sources."
• 3. "Meta-analyses, textbooks, and scholarly review articles are preferred when available, so as to provide proper context."

Eventually I would like for these guidelines to describe textbooks in more depth, since they are so commonly used(and misused), but for now I would just like to propose that "textbooks" in lines 2 and 3 be replaced with "widely recognised standard textbooks written by experts in a field", it is wordy but in my opinion important enough to justify any loss of flow to the prose.AioftheStorm (talk) 17:21, 7 April 2014 (UTC)

I question #2. Academic and peer-review research are notoriously unreliable. The Economist has a lead story [1] about the problem [2]. While college textbooks are far from perfect, they tend to report that which survives the first round (academic journals) and is seen as worthy of teaching to the next generation. Jason from nyc (talk) 22:41, 7 April 2014 (UTC)

The types of mistakes made are on a whole different magnitude. Original research articles are replete with errors in statistical analysis, poor experimental design, and results that cannot be replicated. This is why we do not like original research articles, and instead want review articles which look throughout the literature, seeing what has been able to be replicated, and synthesizing and summarising the knowledge on a topic which has stood for years. Textbooks on the other hand, commonly do no research, propagate popular myths, contain basic and even bizarre factual errors, and are commonly written by people with no expertise in the field. Examples of mistakes include:presenting Dr. Frankenstein's experiment to bring back a dead human being as a real experiment, claiming that "thousands of black soldiers fought for the South during the Civil War", or showing a map of the Earth with the equator running through Texas and Florida. Textbooks will commonly contain theories that were debunked decades ago(like the theory that an asteroid hit the Earth and caused the Moon to form), or present common misconceptions as fact. A description such as "far from perfect" is inadequate to express how unreliable textbooks are.AioftheStorm (talk) 05:07, 8 April 2014 (UTC)

Textbooks are tertiary sources, yes, so they should be removed from that sentence. At the same time, they are pretty reliable, just like other encyclopedias. Whether they are more or less reliable than research papers is really hard to say. My gut feeling is that yes, but only because they stick to simpler and less controversial content, whereas research papers are on the literal cutting edges of their fields. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 09:48, 8 April 2014 (UTC)

I'm amused that while discussing the need for more reliable sources on the Identifying reliable sources talk page, you are taking the word of "DegreeScout.com" on the Frankenstein textbook. --GRuban (talk) 18:32, 8 April 2014 (UTC)

Nah, I disagree with removing "textbooks." Maybe if you demonstrate how they generally are not secondary sources or are generally unreliable, I'll change my mind. Flyer22 (talk) 18:39, 8 April 2014 (UTC)

It depends on your Platonic ideal of a "textbook". Textbooks for children are usually tertiary sources. Textbooks for medical students are almost always secondary sources.

The bigger issue is that to determine whether a textbook is "reliable", you need to compare it to a specific statement that you want to support. A textbook for seven year olds is perfectly reliable for determining that Abraham Lincoln was the 16th president of the US. And you also need to compare it to the typical source: given a choice between a website and a textbook aimed at older teens or young adults, which would you want to recommend? It's all very well and good to say that people ought to spend years studying the scholarly literature, but in practice, the choice for nearly all of our editors is between what they can find on the web and what they can find in school books that they already have in their possession. WhatamIdoing (talk) 00:07, 9 April 2014 (UTC)

I ask not for an unqualified statement of disapproval concerning textbooks, but a rewording of a statement that says textbooks, amongst other sources, "are usually the most reliable sources." Perhaps it is different elsewhere, but most K-12 textbooks in the US are written by a committee of non-experts, reviewed by non-field-specific "education experts", and sent to schools. There usually aren't sources listed for factoids presented, and no real fact checking. Here is a CNN article describing scores of errors from physics textbooks including "Chemistry formulas and physics laws that are so "simplified" they are completely wrong." I'm afraid I have been blindsided by this discussion, I did not imagine that people would contend that textbooks are more reliable than review articles written by experts on a very narrow topic, and reviewed by experts before being published and subjected to prompt retractions if mistakes are noticed from the readership. General purpose textbooks are very lengthy, and they get only a cursory review by people unfamiliar with the subject, as opposed to review articles only a few pages long that get read and re-read and fact-checked by people familiar with the subject. There has been here a misconception that problems with original research articles are indemnable to review articles, but they are two very different types of sources and should not be compared. At a very minimum we should not be recommending textbooks as amongst the best types of sources, and if we do we should qualify it to exclude K-12 textbooks. Also, in response to the above question, there are almost always websites available that are more reliable than textbooks, with the exception being certain college-level and post-college level textbooks.AioftheStorm (talk) 00:54, 9 April 2014 (UTC)

Sure, I'm willing to believe that a good one percent or so of physics websites will be better than a university physics textbook. The other 99% are probably not. By contrast, probably 99% of what's in that high school or university textbook is accurate (which may be the same book: high school textbooks in AP or honors classes are often identical to the first-year textbooks at university). If an editor is picking a source at random, would you rather that he took his chances on a random textbook, which at least has a proper publisher behind it or on a random website, which may be put up by some quack?

Nobody is really claiming that review articles are necessarily better than textbooks. What we're saying is that the textbook is more accessible to editors, more uniform in quality, and perfectly appropriate for most of the basic statements in an encyclopedia article. WhatamIdoing (talk) 02:50, 9 April 2014 (UTC)

I'm willing to accept the above post. I believe the mention of textbooks under secondary sources should be removed, however, since if you're using a textbook at the university/grad/post-grad level that is a secondary source then you should already be knowledgeable enough to know what a secondary and tertiary source is, and if you don't know how to tell the difference between a secondary and tertiary source, then you're probably at an entry university or lower level in education, in which case all the textbooks you would be using would be tertiary sources. Textbooks are generally described as tertiary sources anyways, much like encyclopedias even though specialty encyclopedias which are secondary sources do exist.AioftheStorm (talk) 05:15, 9 April 2014 (UTC)

Most grad students outside of history don't know what a secondary source is—and even in the field that originated the concept, some experts can't agree on how to classify some sources. Some of the Wikipedians who wrote these policies had no idea what the term meant. I just spent a couple of years trying to disentangle "independent" from "secondary". I think that we've cleared it all up by now (please squawk if you find problems), except for the WP:Notability guideline. Simple statements like WP:Secondary does not mean independent are news to some people. The fact that routine news stories are primary sources is news to most editors (this is why fixing the language in the GNG is impossible). We really should not assume that anyone has any idea at all.

On the practical point: textbooks, even if aimed at 12 year olds, are often better sources than the alternatives. Putting textbooks in the class of usually good sources does not mean that, for any given statement, a non-textbook source might be better.

Finally, if you haven't read WP:MEDRS, then you'll like it. WhatamIdoing (talk) 20:55, 9 April 2014 (UTC)

• WP is much too focussed on "secondary" and (in typical WP style) this has arisen because it's simple to judge, not because it's useful.

We favour secondary sources, not because two is a magic number, but because "secondary" is hopefully a point where accuracy and objectivity are at a useful maximum. Yet WP then turns this into avoiding court records as being primary – Why? Are we afraid of biased stenographers? We are now turning against textbooks, because if they're tertiary, they're seemingly too far from the truth. Both of these are nonsense and should be rejected. We do not favour secondary sources because they are secondary, we favour them because they will in general offer a concrete advantage to our real goals of verifiable and accurate detail. In the cases where this isn't so clearly bound to this secondary nature, then we should follow the advantage, not the dogma. Andy Dingley (talk) 10:09, 9 April 2014 (UTC)

No, we're afraid of biased speakers in the court, and of biased editors cherry-picking statements out of them. WhatamIdoing (talk) 20:55, 9 April 2014 (UTC)

So what? A biased speaker in court is still something that (recorded verbatim) we would regard as evidence of their words to that court. It's a source that they spoke those words in the court and also that they expressed a particular opinion. If that opinion is inaccurate or biased, that's a matter of their inaccuracy, not the inaccuracy of a stenographer. We should use such a source when reporting about them personally, just as we might with a secondary newspaper source, but be wary of both if treating the speaker as an authoritative source of facts.

Cherry-picking by editors is an issue for any source, reliable or not. Andy Dingley (talk) 13:40, 10 April 2014 (UTC)

We worry about this because we have had serious problems with this. A court transcript is not a good source for supporting a statement like "John Doe 'is definitely a murderer', according to court documents", but they have been used this way. We also have had a problem with privacy invasions (e.g., birth dates, home addresses, and the names of family members). If no secondary source sees fit to mention these, then they don't belong in an article. The community banned these as sources because they weren't "offering a concrete advantage". WhatamIdoing (talk) 20:54, 11 April 2014 (UTC)

Clear policy statements?

Can someone please point me to the WP:POLICY that supports the following:

1. Only WP:RS are allowed in articles. Non-RS, and the content they support, may be removed without discussion.
2. WP:3RR does not apply (as it's exempted for some BLP issues) in the case above.

I know of no such policy. We have policies that some statements must be supported by RS, including the get-out clause that any content may be required to have RS in support of it if the challenger holds a finger in the air and turns three times widdershins, reciting the mantra, "I challenge thee". However I still know of no reason that non-RS should be removed if they are being used in addition to any other sources, and that there are sufficient RS to support the content. Andy Dingley (talk) 14:01, 8 April 2014 (UTC)

The policy in question is: WP:Verifiability. More specifically the WP:BURDEN section of that policy. Blueboar (talk) 14:15, 8 April 2014 (UTC)

The agreed upon policy, as I understand it, is that all material must be verifiable in a reliable source. This does not prohibit providing a reliable source that is sufficient to support the material, and also providing unreliable sources that support the same material. Jc3s5h (talk) 15:02, 8 April 2014 (UTC)

I would think consensus would normally support removing non-RS from an article, as it's by definition not reliable, and therefore makes the article appear unreliable, but perhaps you are using non-RS to mean something else, as you mention BLP, perhaps you are referring to self-published -- those can be reliable, under WP:BLPSPS. Alanscottwalker (talk) 16:40, 8 April 2014 (UTC)

I'm not looking for consensus, but for WP:POLICY in particular.

As to the justifiability of removing them, then that's going to depend on context: how far are they from RS, how useful is the content they contain, how credible is it (in particular for this item, not for judging the source overall) and most importantly, what is the content they relate to in the article. Our rules on RS are quite strict - especially against forum sources. So we (rightly) can't rely on these, yet some of these can still be valuable additions to the article. I don't know of any policy against them, I've certainly seen non-RS sources that improve the encyclopedia without compromising WP:V – we have to judge case-by-case. Andy Dingley (talk) 22:40, 8 April 2014 (UTC)

I can't imagine a case where an unreliable source improves the encyclopedia. They are, by definition, unreliable. In some cases we may use primary sources, or non-independent sources, but unreliable sources are of no value to the project as sources. If content can only be supported by an unreliable source, WP:V permits us to remove it. Pburka (talk) 22:49, 8 April 2014 (UTC)

No-one is questioning the case when content must rely on RS. However there is still no policy against sources that don't meet RS (which doesn't mean they're unreliable, RS is very picky) when used supplementally to this. Andy Dingley (talk) 23:00, 8 April 2014 (UTC)

Are you proposing some sort of source which exists between reliable sources and unreliable sources? By my logic, any source which is not reliable is unreliable. Could you perhaps provide some examples of such sources so I can better understand your argument? Or are you saying that a fact which has been referenced to a reliable source might also have references to unreliable sources? In that case, the unreliable sources still serve no purpose. At best, they're harmless and at worst link spam. Pburka (talk) 01:16, 9 April 2014 (UTC)

I'm not proposing anything – I'm trying to find the supposed WP:POLICY (in caps) that is a blanket ban on any use of non-RS whatsoever. Can you cite it?

As an example, [3] shows that the Canadian Ford Lynx scout car ended up a foot taller than the original Daimler Dingo, owing to a technical issue of suspension design (I'm writing an article on this). As it has "forums" in the URL, it blanket fails WP:RS. Yet it conveniently photographically illustrates the extra height. The Army-published handbooks just give two figures (59" and 70"), which are nothing like as obviously convincing. Andy Dingley (talk) 02:02, 9 April 2014 (UTC)

User:Blueboar already pointed you at the appropriate policy. WP:V is a very clear policy. All material must come from a reliable source. Forums aren't reliable sources, so don't cite them. Cite a reliable source, or don't include the material at all. Pburka (talk) 02:28, 9 April 2014 (UTC)

Yes, WP:V very clearly states nothing of relevance to this question! WP:V states that challenged content must be supported by RS. It states nothing about any supplemental sources beyond this. Andy Dingley (talk) 03:14, 9 April 2014 (UTC)

For non-BLP disputes, 3RR definitely applies. It doesn't matter if the person is citing the Time Cube website, or even no source at all: you simply may not edit war to remove (non-BLP) material.

Depending on the article, you might want to convert the non-RS sources to WP:External links under the WP:ELMAYBE clause that specifically permits links to unreliable sources. WhatamIdoing (talk) 00:11, 9 April 2014 (UTC)

Just for Pburka, here, just off the top of my head, are at least three very common cases where clearly unreliable sources noticeably improve the encyclopedia.

• a photo of Jane FamousPerson, taken by me, and released under an appropriate license. Clearly unreliable, yet a major help to the article.
• a link to a blatantly-unreliable-tabloid article exposing a BLP scandal (that was then picked up by many more reliable sources; but the tabloid was the original).
• a fraud or a hoax (that became notable in itself).

I'm sure there are plenty of others. In the latter two cases, we require reliable sources in addition, but we don't delete the unreliable ones. --GRuban (talk) 16:02, 9 April 2014 (UTC)

1)True - one of the more intriguing conundrums on the pedia (to me anyway).

2 and 3 - be aware of WP:BLPEL. So, sometimes we do delete, depending on content. -- Alanscottwalker (talk) 16:34, 9 April 2014 (UTC)

None of those are suitable as references. You can use an image to illustrate an article. The tabloid article might be a valid citation as a primary source, i.e. if one were quoting the article. I have no idea what you mean about the hoax: an event isn't a source. Perhaps I've misunderstood your question, and you want to add unreliable sources in some context other than as references which you've not described. Pburka (talk) 18:14, 9 April 2014 (UTC)

The third is; it's a primary source that can be used as a source for what it actually said. We actually prefer the original hoax when we are sourcing direct quotations. WhatamIdoing (talk) 20:56, 9 April 2014 (UTC)

A hoax is an event or a concept. If the hoax originally appeared in a publication it may be appropriate to cite that publication as a WP:primary source, and it would be a reliable source in that particular context. Note that it is the publication, not the concept of the hoax, which we use as a source. Crop circles, for example, are hoaxes, but are not suitable as references. Pburka (talk) 23:37, 9 April 2014 (UTC)

You wrote "All material must come from a reliable source." That is not so. In each case I cite, it doesn't come from a reliable source, by our common definition. And yet it is definitely valuable Wikipedia material. This material might be called "illustration" or "primary source" or "not reference", but that does not make it not "material". So the answer to the question Andy Dingley is asking, is, yes, there are times when we do want to include material that does not come from a reliable source, and they're fairly common. --GRuban (talk) 19:18, 10 April 2014 (UTC)

This is so, because "reliable source" is defined as being dependent on the context. A source that is not usually reliable for most purposes might be the most reliable and authoritative source possible for another purpose. WhatamIdoing (talk) 20:56, 11 April 2014 (UTC)

Vox

Is this new voice RS or an unreliable blog? Thanks. Lfstevens (talk) 04:35, 10 April 2014 (UTC)

Lfstevens... you should probably ask this at WP:RSN (the Reliable Sources Noticeboard). This talk page is really for discussing the wording of the guideline itself.

That said... I can all but guarantee that the first question they will ask you is: "Reliable for what exactly?" You see, context matters. The same source may be absolutely reliable in one context, and completely unreliable in another. For example... a pamphlet published by the Flat Earth Society would be reliable for the statement "The Flat Earth Society contends that the Earth is flat"... but it would be completely unreliable for the non-hedged statement "The Earth is flat".

We also need to look at the appropriateness of the statement itself... in the context of the article topic. For example, the opinion of the Flat Earth Society about the shape of the Earth is appropriate to mention in the context of the article on the Flat Earth Society itself... but their opinion would not be appropriate to mention in the context of our article on Earth (for more on this, see our policy on WP:UNDUE WEIGHT).

In other words... reliability depends on exactly what you want to say... and where you want to say it. Blueboar (talk) 13:11, 10 April 2014 (UTC)

Dear sir I am new on wikipedia and I wanted to write correct thing... pl  help me to do...--SjElle (talk) 15:59, 11 April 2014 (UTC)

Lapia

Lapia is a town in niger state nigeria — Preceding unsigned comment added by Nett nitter (talk • contribs) 15:09, 14 April 2014 (UTC)

Statements of opinion

This is unhelpful, since it is circular: "Some sources may be considered reliable for statements as to their author's opinion, but not for statements asserted as fact ... A prime example of this are opinion pieces in mainstream newspapers."

Naturally, opinion pieces are statements of opinion. That doesn't help me figure out what to treat as opinion and what to treat as fact. I think there should be some wording to the effect that any statement which is subjective (by community consensus, perhaps) should be treated as a statement of opinion. Howunusual (talk) 22:16, 14 April 2014 (UTC)

Unfortunately, I think it's one of those "if you need to be told, then you won't understand" problems. It's obvious to most people that a statement like "the sonnet is the best type of poem" is an opinion, and that "sonnets rhyme" is a statement of fact. But you should not use a purely opinion-oriented piece to source a statement of fact, even though the opinion piece might be completely accurate in its facts (because it might not be accurate).

Also, while there are clear examples, there are also a lot of statements that are difficult to classify. Is "the economy is down because of this policy" a fact or an opinion? People won't agree on this. WhatamIdoing (talk) 03:14, 15 April 2014 (UTC)

If experts don't agrree, as in the case of affects on the economy, then it should be treated as an opinion. If it can't tested in some scientific sense, then it should also be treated as an opinion. For example, "war is bad" is probably very non-controversial, but it's still an opinion. Wikipedia has a pretty clear policy about the first case. It doesn't seem to acknowledge the idea of an inherently subjective belief, at least on this page. I can't really speak to how it's usually handled by the community. I just don't see a clear policy about it here. Howunusual (talk) 18:30, 16 April 2014 (UTC)

It isn't just subjective statements... some "facts" are controversial, not fully accepted by the experts in a particular field. In such cases, the "fact" needs to be presented in Wikipedia as the opinion of those who believe it is (in fact) a fact. We do this through attribution. Suppose you read a popular book by a conspiracy theorist that stated definitively that the US Government is controlled by the Illuminati. We can not write "The US Government is controlled by the Illuminati <cite book>"... A conspiracy theorist is not a reliable source for such a definitive statement. However, if we attribute the statement, and write: "According to Ima Nutcase, a popular conspiracy theory author, the US Government is controlled by the Illuminati <cite book>" the situation changes... We are no longer making a statement about the government... we are presenting a statement about Ima Nutcase's opinion. That statement is verifiable and reliably sourced to his book. (note... this isn't the end of the story... that this statement of opinion can be reliably sourced does not guarantee inclusion of that opinion in a particular article... see WP:UNDUE). Blueboar (talk) 11:29, 15 April 2014 (UTC)

Hi. I agree with what you said about controverisal facrts. But, I was wondering about guidelines for subjective opinions. Even "consensus" opinions should be represented as such. Howunusual (talk) 18:33, 16 April 2014 (UTC)

Really, really bad English

Maybe this is a small point, but I recently bought a used souvenir book on a Spanish cathedral. The book, apparently translated from Spanish, regularly calls the cathedral a "he," opening up OR/SYNTH problems in using it. And I lknow of at least one reference book from Academic Press, written by an English-as-an-at-least-third-language academic which at several points is basically incoherent. Does a source have to be at least clearly comprehensible to be reliable? John Carter (talk) 15:26, 25 April 2014 (UTC)

A lack of comprehensibility would likely imply questionable reliability, I'd have thought. As always though, one would have to look at the specifics. Of course, it isn't unknown for academics (even respected ones) to be incoherent in their first language... AndyTheGrump (talk) 15:36, 25 April 2014 (UTC)

Poor translation doesn't mean that the author knows nothing about the subject. You could probably figure out what the pronouns refer to, for example. However, poor translation might also make it impossible for you to get the author's knowledge out of the text. Some things (names? dates?) might be easy enough to sort out, but other passages might lead you astray. IMO it would be preferable to look for another source when you think that another is reasonably available. Looking for the original, non-English source might also be a good approach. WhatamIdoing (talk) 23:17, 26 April 2014 (UTC)

I agree... It does not really matter whether a poorly written source is reliable or not... since there is a high probability that any information worth gleaning from the poorly written source will also be found in other sources that are even more reliable. Blueboar (talk) 13:16, 27 April 2014 (UTC)

should this change be made?

At Talk:List of Bohemian Club members there is an ongoing discussion about whether a person listed as an honorary member in an SPS from the club thus is automatically an honorary member of the club. And whether a person citing such an SPS source then automatically validates the claim that a person is an "honorary member" because once a secondary source refers to the first source, it becomes "fact."

My belief is that if the original source is questionable, then any cite of a questionable source still does not meet WP:RS. (E.g. the NYT has many times now written about hoaxes as though they were fact - but that did not make the hoaxes into facts.) If any editors here have opinions on that issue, the discussion is perpetual there.

In which case we need to add:

Any organization listing of members, even in a self-published source, is automatically a reliable source for ascribing membership to such people named.'

Ought we do so? Collect (talk) 14:21, 30 April 2014 (UTC)

Assuming the original actually is from the Bohemian Club, then it is not a questionable source... it is an appropriate use of a reliable primary source (You can not get a more reliable source for a statement about the membership of an organization than a published statement from the organization itself). Of course, we would have to be sure that the published statement actually is from the organization (and not a fraud... which, given all the conspiracy theories that involve the Bohemian Club, we have to consider as a real possibility). Blueboar (talk) 15:03, 30 April 2014 (UTC)