Wikipedia talk:Verifiability

Lets stop selling anecdotal evidence as "truth" - show the beef, please

I want to be shown a long lasting (a year or so) consensus (ie a result not of under-attention, but of willful misintepretation of policy by a set of editors) edit in which the sources are contradictory and there is counterfactual information shown as factual information. I do not care about the wikilawyering POV warriors and Fringemongers do in the talk pages - people constantly argue even against explicit policy if they believe they can get editors convinced -, I am talking about actual, live, article content that is not factual but is presented as factual. I have yet to see one such example. When those edits happen they get reverted, and rather quickly. If shown one, I will change my position on "v, not truth".--Cerejota (talk) 16:27, 15 September 2011 (UTC)

OK, in the article on Frederic Chopin: "The French passport was issued on 1 August 1835, after Chopin had become a French citizen." This is based on what one source says (about his "changing his citizenship"); no other source has been found that mentions any change of citizenship (as we would expect them to if it happened), but (a form of) this statement was forced into the article by an editor saying "we have to go with what the source says". Hence a prominent Wikipedia article is probably telling the world an untruth, and has been for - what - well over a year, I think.--Kotniski (talk) 16:37, 15 September 2011 (UTC)

(edit conflict)But the truth is he did become a French citizen!--Cerejota (talk) 17:24, 15 September 2011 (UTC)

No it isn't. Chopin's father was French, which makes him a French citizen from birth. See footnote 18 in the French article, which is based on the same source (Szulc) that appears to have been used to put forth the claim that he became French. Szulc stated the facts, and then drew a conclusion that appears to be incorrect. If this conclusion was not widely reported in reliable sources, it is dubious and we should not report it either. Hans Adler 07:29, 16 September 2011 (UTC)

Yes, I should perhaps have said that the statement is (probably) literally true (if he became a French citizen at birth, then he did get his passport "after" (a long time after) becoming a French citizen), which is probably why we gave up trying to take it out, but its obvious apparent meaning - that he obtained citizenship fairly immediately before getting the passport - is (probably) false.--Kotniski (talk) 07:43, 16 September 2011 (UTC)

Cerejota, there are examples, and in fact many are examples I, and many others who are worried about the "not truth" wording, think should occur. For example every time we do NOT put in the latest news on a scientific subject because we are waiting for a proper publication. That sort of "wrongness" is ok. Remember that people worried about the not truth bit are not asking for "yes truth" either. We just want existing policy explained clearly. But I would like to raise a couple of points about the way you set your challenge up:

• First you exclude examples which involve "POV warriors and Fringemongers". But because there is no sharp definition about who these are (they never self identify) it means you could argue, if you wanted to, that by definition any example is disqualified.
• You demand to see live article content. Why? Not only can things which are live examples be acceptable, but I think most people would agree that things which do not survive as live content can still be problems for Wikipedia's normal operation. They can slow down or stop improvement of controversial articles. And they can also chase away new editors. Do we ignore these problems?
• A point I made elsewhere about "truth": when a well informed newbie arrives and says an article is not up to date, because he knows of new information on a subject which is not yet published, should we chase him away and tell him we do not care about truth on Wikipedia, or should we say "Great, but do you know when it will be published?" The answer our community chooses will have an impact on the quality of the result. It sets a tone.

It seems a bit like a challenge that will take discussion nowhere?--Andrew Lancaster (talk) 17:16, 15 September 2011 (UTC)

But the truth is he did become a French citizen!--Cerejota (talk) 17:24, 15 September 2011 (UTC)

(edit conflict)And constantly claiming that "verifiability, not truth" is a problem without concrete evidence of this being a problem does move the discussion forward? My challenge is not out of the blue, it is a direct response to the claim that there is a problem that quite frankly doesn't exist - because not a single example of the problem in the wild is shown. Just one. I completely disagree with your second critique: WP:BRD is how the joint works, has worked, and should continue to work - the process of reversion of good faith edits is as important as the process of adding information. Your third point is thoughtful and thought inducing, and the kind of argument I am amiable to, even if I am not convinced at this time of its validity - I do agree however that the impact on new users of how policy is worded is an important consideration. --Cerejota (talk) 17:42, 15 September 2011 (UTC)

• Well, this won't be productive. The pro-change party run round finding examples, and the anti-change party quibble them. Didn't we do this for about eight months before?—S Marshall T/C 17:29, 15 September 2011 (UTC)

That is only how long you have been involved in it. I was an editor (IP I think, but maybe already account) when the phrase was added to the first sentence. It has never been completely accepted. However, V is the most successful of policies, more successful than NPOV even, and perhaps second only to BLP. And it has been a success with the damn sentence. If ain't broke, don't fix it.--Cerejota (talk) 17:47, 15 September 2011 (UTC)

Cerejota, both sides of the dialogue agree V is important. It is a question of whether we can improve the tone of the exact wording. Saying, as we hear time and time again (as in your edsumm) "if it ain't broke don't fix it" begs a question of whether this folk wisdom is always automatically correct. Of course everyone knows it is not. I think a similar argument confronted Galileo? Please also consider what you would think if you were being continually challenged to show a concrete live example of the two words "not truth" improving an article. It is an example of rhetorical skill to deflect discussion in this way, but....--Andrew Lancaster (talk) 18:36, 15 September 2011 (UTC)

I think we need a Godwin's law for Galileo. :) I can provide you a clear example: September 11 attacks. "V not truth" has kept truther conspiranoia from tainting that article. And I have tons of other examples. Even here, ITT, the Chopin example: it is clear that Chopin became a French citizen, and this is well sourced, yet an editor, ITT, claims this not to be true. Well, I answer: verifiability, not truth. Find me the reliable sources who say Chopin didn't become a French citizen. You can't, because they do not exist. And so on. --Cerejota (talk) 04:57, 16 September 2011 (UTC)

Sorry, what are you talking about? It is sourced that he became a French citizen, but it is not clear at all that he became a French citizen (unless you have some more information to contribute, which would be very welcome). You seem to be demonstrating the exact problem that you claim doesn't exist - because we can't find sources that say "he didn't become a French citizen", and there is one source that implies that he did, there are editors (including apparently you) who think that our work as encyclopedists is done, and that we should unthinkingly include the (probably inaccurate) information in the article, to the obvious detriment of Wikipedia's usefulness as a source of accurate information. Of course we can't show a direct causal link between the wording "verifiability not truth" and this kind of irrational editor behaviour, but - well, you said if you got an example you would change your view on that matter, and you've now got an example, so I assume your view has indeed changed.--Kotniski (talk) 06:59, 16 September 2011 (UTC)

And as to the September 11, it isn't "verifiability not truth" or any other policy statement that has kept conspiracy theories out of the article, but editors who know what WP's principles are and are prepared to "fight" to uphold them. If they are helped in doing that by the ability to quote policy statements that describe those principles, then they would be helped just as much (or even more) if the relevant policy statements described those principles accurately.--Kotniski (talk) 07:07, 16 September 2011 (UTC)

Exactly. When we are keeping fringe theories or poorly sourced content out of an article, what we are doing is evaluating for accuracy and authority. This has nothing to do with "V not truth", and everything to do with evaluating sources. The only thing "V not truth" has achieved, is that it has confused editors by falsely attributing an equivalency between all sources and their respective content. Many editors read "V not truth" and assume that it if it can be sourced, it can/should be added. That this is still in the policy is unbelievable. Viriditas (talk) 10:29, 17 September 2011 (UTC)

This is degenerating back into tossing and parrying talking points. Cerejota, I feel that there are several flaws in what you just said. But I don't intend to go back there. We are working on a compromise. North8000 (talk) 18:49, 15 September 2011 (UTC)

• Yeah. We went this way before. Evidence gets produced, and quibbled. More evidence gets produced, and quibbled. The way to deal with editors who refuse to be convinced is to disregard them and seek to convince their audience instead.—S Marshall T/C 19:39, 15 September 2011 (UTC)

Actually I've followed the discussion for quite a while, and your implication that evidence was provided "for about 8 months" is false, as far as I know. Evidence was requested for 8 months, but only in the last month has any real evidence been provided. To "quibble", I'd say it's of varying success in demonstrating a problem, but some of it is interesting and relevant. Also although you have been beating this drum since near the beginning, if you ever provided any evidence yourself then I didn't see it. Quale (talk) 00:25, 16 September 2011 (UTC)

--Cerejota (talk) 04:57, 16 September 2011 (UTC)

The point is that any example you can bring will be just as controversial or uncontroversial as examples of the opposite. We already know this can go back and forth forever. But for the reasons I gave above, in the end such examples are simply not a convincing argument that we should not try to improve the wording of a policy. There is no logical connection. In other words it is a red herring, not part of any possible constructive conversation, but rather aimed at trying to block any discussion about whether the wording can be improved. In the end this approach comes down to the "if it ain't broke" excuse for any status quo.--Andrew Lancaster (talk) 05:54, 16 September 2011 (UTC)

• Quale, the reason I don't run around producing evidence on demand is that I've been editing Wikipedia for a long time and I've learned some things about editor behaviour. (I learned quite a bit more from this page...)

  A balanced picture of the discussion is that while I produced only one example of evidence in support of my view, I also asked for evidence from others in support of retaining "not truth" for a similarly long period, without any result at all. It took until a couple of weeks ago for your side to produce any evidence at all—and, in stark contrast to the behaviour of others in this debate, the moment your side did produce evidence, I acknowledged it and changed my position.—S Marshall T/C 07:40, 16 September 2011 (UTC)

I've responded with the requested examples of the current problems many many times, as have others. Except for a few cases where someone does this User:North8000/Page2 the thread typically just ends when the example is given. Then, shortly thereafter, someone again says "give an example" or "nobody gives examples".  :-) North8000 (talk) 11:27, 16 September 2011 (UTC)

Just an aside on the Chopin debate... the factual debate isn't really about whether he became a French citizen... the factual debate is about when he became a French citizen (at birth, or at some later point?). The sources are inconclusive. The point is... its ok for sources to be inconclusive and contradict each other. Its why we have WP:NPOV... When reliable sources disagree we present both sides of the debate... phrasing the different viewpoints as being the opinions of proponents of each side, and not stating them as fact. Blueboar (talk) 12:34, 16 September 2011 (UTC)

Unfortunately in this case it's not quite like that (at least, based on the sources we had at the time - of course, if someone's found some more since then, the whole question might be moot). There was no reliable source that said he became a French citizen at birth, or anything else that would directly contradict the claim that he changed his citizenship to French later. There was one source that said he changed his citizenship. So there was no contradiction between sources, no factual debate to report... The contradiction was between what that one source said, and the conclusion that intelligent Wikipedia editors came to based on various evidence that I could go into if you wanted (including the fact that no other source mentioned such a change). According to the "verifiability not truth" crowd, we don't care about the other evidence, we just parrot the one source and think we've done the world a service.--Kotniski (talk) 13:58, 16 September 2011 (UTC)

It is one source in the article, many sources otherwise. However, and again, why you say this is not "true"? Here is the thing people are missing: truth in wikipedia is determined by the careful evaluation of sourcing, NPOV, and generally known as facts. We have different policies for each of those things: WP:NPOV, WP:V, and WP:FRINGE. I consider the position that Chopin didn't become a french citizen to be WP:FRINGE.--Cerejota (talk) 18:12, 16 September 2011 (UTC)

Well, as long as we are getting into it, there are basically two things when talking about accuracy (or the accuracy-related meanings of "truth"):

1. The framework of the answer, and is it at least 99% agreed on? For example, the framework of "who won the 2001 USA World Series of Baseball" is 100% agreed on (per the rules of the governing body) The framework of "is Pluto a planet" is maybe 90% agreed on (the word of the most accepted body) and the framework of "is Obama a good president" (e.g. what does "good" mean?) is nowhere near 99% agreed on .
2. Per that framework, to what certainty is the answer known per normal accepted objective methods?

And occasionally, you have chaos until you precisely define the question, e.g. with global warming, after which you can start on the above. 18:52, 16 September 2011 (UTC)

Replying to Cerejota - yes "truth in Wikipedia" (well, I thought you were against the concept of "truth", but I know what you mean) is indeed determined by the careful evaluation of sources. What was done in this case was not careful evaluation at all; it was "one source says X; no source directly says NOT X, therefore X is true". This kind of careless reasoning is, it seems to me, likely to be encouraged by wording like "verifiability not truth" and "the threshold", which imply that once a source has says something, we are compelled to switch off our critical faculties and unquestioningly include that information in Wikipedia.--Kotniski (talk) 16:58, 17 September 2011 (UTC)

A long time ago when I was active on some Israel/Palestine related aticles, there was this dispute on the Hamas page about a sentence in the lead, saying something along the lines of "Hamas is best known for terrorist attacks" and that cited to a reliable source. That sentence was the focus of editing disputes; but at the time I was editing there, you could't get it removed and replaced by a more factual statement. It had stayed in the article for a very long time.

The problem was that any argument was shot down because that statement was in a "reliable source" and therefore protected by Wiki-policy. You couldn't argue that the source, a book about the Mid East, makes the statement as just an opinion of the author which isn't intended to be taken to be taken serious a reliable scientific statement. That would be "OR" and my attempt to remove that statement and replacing it with something along the lines of "Hamas has committed terror attacks" with sources led to me getting a warning by an Admin (I think Humus Sapiens) on my talk page.

I can only guess why particularly the pro-Israeli editors were feeling so strongly about keeping that non-factual sentence. My guess is that they wanted to stick to using Wiki-policies as a tool to protect what you couldn't otherwise protect. So, if they were to give ground on this issue, even if that meant an improvement from their perspective on that particular point, they could no longer invoke Wiki policies on other points in other articles in this way.

So, there were estabished editors who were active WIki-editors for many years and quite a few Admins, who were behaving in that way for years here. After I left, there have been one or more ArbCom cases, and I think things have changed, but I haven't followed the I/P related articles and thse ArbCOm cases. But I don't think ArbCom has ever made an issue about editors following Wiki-policies being a problem, so they may not have identified all of the relevant problems.

Another area that I have followed closely for years are some climate change related articles. There there was pressure from climate sceptical editors to include unreliable opinions from sources. They didn't succeed, but the reason why they didn't succeed was, as seen by ArbCom, due to disruptive elements (note that while William was perhaps always a bit controversial in the way he acted, other editors like KimDabelsteinPetersen were also topic banned). Count Iblis (talk) 15:25, 17 September 2011 (UTC)

But what did the sources say Hamas was best known for at the time? If the sources generally agreed that Hamas was best known for terrorist attacks (as opposed to, say, their many charitable endeavors, and now their many difficulties with running a government), then it's not unreasonable for us to report this. What Hamas is best known for among certain subsets of people might be different (and a very interesting subject), but I suspect that the statement was actually factual, at least to the extent that you were considering only the statements from English-language sources about what Hamas was best known for at that time. These results "best+known+for"+-wiki+-wikipedia suggest that the sources might have supported the claim that you have labeled as false. WhatamIdoing (talk) 04:27, 18 September 2011 (UTC)

Yet another example of a situation where a simple rewrite would have resolved the issue. Statements about what a subject is "best known for" are, by definition, opinion, rather than fact. So the sentence should have been attributed... along the lines of "According to X, Hamas is best known for terrorist attacks <cite X>". This would make the statement a) verifiable, b) not OR, and c) presented with a neutral point of view. Blueboar (talk) 12:51, 18 September 2011 (UTC)

I would say it's something of a grey area. I don't think "Neil Armstrong is best known for being the first man on the moon" would be regarded as pure opinion, and to attribute that statement to a source in-text might well be regarded as a violation of NPOV, as implying some doubt where there is none. (Of course, it's not normally necessary to make such statements - you just say appropriately early in the lead that Armstrong was first on the moon or that Hamas do terrorism.)--Kotniski (talk) 08:05, 19 September 2011 (UTC)

I think what would make the Hamas case different is that there was no consensus. When there is no consensus, we should err, I believe on the side of inclusion and attribution, which is also apparently what Blueboar is saying. But where does WP:V tell people this? If I understand correctly Count Iblis is saying that WP:V was successfully used as an excuse to avoid this correct approach. If so, then perhaps this was another example of what I have been calling the imperialism of the current WP:V wording, whereby it seems to over-ride the other core content policies.--Andrew Lancaster (talk) 08:16, 19 September 2011 (UTC)

Do you mean consensus between editors, or consensus between sources? If the former, it's hard to speak of consensus when the discussion is flooded with people with a point of view to push. If the latter, then the problem is often that one source makes a statement that no other source directly contradicts (e.g. we might not have a source saying explicitly "Hamas is best known for its charity and educational work"), but nevertheless appears not to be supported sufficiently by the body of sources as a whole for us to state it as fact (or even report it at all).--Kotniski (talk) 08:22, 19 September 2011 (UTC)

I meant consensus between editors, and I recognize the two problems you mention. No policy will ever stop such problems completely, but the normal practice when everyone is working according to best practice and good faith is that strong conclusions such as "there is a consensus" or "Z is most famous for X" should either be uncontroversial, like with Niel Armstrong, or else there should be editor compromise by using things like attribution, and inclusion of balancing context.--Andrew Lancaster (talk) 08:41, 19 September 2011 (UTC)

And back to the subject, I think the concern is that WP:V does not give any indication that such best practice exists, and it therefore tends to be used by some editors in order to edit in ways which are not best practice.--Andrew Lancaster (talk) 08:43, 19 September 2011 (UTC)

We can't cover every possible area of best practice in WP:V, though. All we can do is try to write a simple policy that makes sense, which acts as a springboard for editors to use their judgment. My position is that WP:V needs to be stripped down and simplified, not added to.—S Marshall T/C 11:07, 19 September 2011 (UTC)

The two aims are not necessarily in conflict with each other.--Andrew Lancaster (talk) 11:44, 19 September 2011 (UTC)

Re Andrew's question: But where does WP:V tell people this? - I think we need to ask a more fundamental question first... Is attribution something that is within the scope of WP:V... or should it be said in some other policy (such as NPOV)? In other words... is this something that should be said in WP:V?

(Note... I am of two minds on this question... which is why I decided to address it in my proposed re-write at WP:V/First sentence) Blueboar (talk) 11:47, 19 September 2011 (UTC)

Well yes something could be mentioned or a cross reference to another policy could be give, but I guess that prior to needing to consider adding in something about it, one point is that WP:V should at least allow room for the fact that certain matters require consideration of more than just WP:V.--Andrew Lancaster (talk) 12:02, 19 September 2011 (UTC)

More fundamentally, we should decide whether WP:V should exist at all. Only once we've discussed rationally whether we really do need two separate pages (V and NOR) will be be in a position to know what the scope of each is supposed to be. (And even if we do have two pages, we ought to consider what their titles ought to be, to reflect the scope we've selected.)--Kotniski (talk) 12:16, 19 September 2011 (UTC)

Looking at WP:V and WP:NOR I agree there is a LOT of redundancy between the two pages.

Much of what is in WP:NOR could be moved to WP:V and WP:NOR streamlined with better examples (WP:SYN's are somewhat poor) and some ideas pulled from WP:NOTOR to clarify things.

For example a WP:SOURCECONFLICT section where examples of how NOT and how to handle RS that conflict would be a welcome addition to the WP:NOR page (as this how fun filled mess shows we need one).--BruceGrubb (talk) 16:46, 19 September 2011 (UTC)

Strong support for "verifiability, not truth," What a citation says can be agreed upon by editors of diverse opinions, "truth," is not so amenable. Other issues of reliability and POV and relevance should be discussed elsewhere, and cross referenced. If some reliable source suggests Chopin was a citizen (if it is relevant), we should report that. If no reliable source says he was not a citizen, perhaps it is irrelevant but that does not effect what a reliable source in fact says.Alanscottwalker (talk) 22:02, 23 September 2011 (UTC)

The flaw in this way of thinking is that, while it may be that editors will nearly always agree on whether a source says something, there is plenty of room for disagreement as to whether a given source (possibly in isolation) is reliable on a given point. We do humanity no service if we switch off all our other critical faculties and just report everything we read in any book, journal or newspaper as if it were unquestionably true. No source is 100% reliable, and we know that. We would be extremely incompetent encyclopedists if we really did adopt that philosophy (fortunately, in practice, we usually don't).--Kotniski (talk) 10:12, 24 September 2011 (UTC)

I do not think Alanscottwalker's thinking is flawed, even though I think I see Kotniski's point. And I think most editors who want to tweak the "not truth" wording agree with his explanation, even if they want "not truth" expanded upon or moved out of the particular position it currently has in the first sentence. The main concern is that not everyone understands the current sentence. @Kotniski, I see what you mean, but in reality I think our policy pages aim to make verifiability a simple criterium for inclusion, whereas concerning "reliability relative to a subject" (which is something very hard to distinguish from "due weight" - watch WP:RSN) we actually want editors not to use simple rules, but rather to discuss the details of the context. For me "verifiability, not truth" is about telling editors not to waste time debating about what is true, but to spend time more constructively debating about what is the right way to weight things and get the wording right on a particular subject. In simple words, the way we avoid getting stuck arguing about what is true is by limiting ourselves to writing about what is published. That is not because we do not care about what is true, but because that is the only practical way we can work. (I am thinking about a recent comment of User:Cerejota here and basically saying that WP:NPOV is what requires WP:V. If WP:NPOV was not so important and difficult, we would not need WP:V.)--Andrew Lancaster (talk) 10:47, 24 September 2011 (UTC)

I agree that it is not Alanscottwalker's thinking that is flawed, IMO it is the premise that the phrase "verifiability, not truth" has a well-understood meaning that is flawed.  And yet by Alanscottwalker's own logic system, we have that editors cannot agree on what is truth, so he/she should readily also agree that editors also will not agree on the meaning of "verifiability, not truth".  Unscintillating (talk) 11:31, 24 September 2011 (UTC)

I certainly agree with that last point; though the same applies not only to "truth", but also to "verifiability" - not only is verifiability a suboptimal word to express a concept that is really more like "sourceability" or "attributability", but also, editors do not agree on any simple definition of this magic concept. To imply that there is some relatively simple "threshold" is in itself misleading, quite independently of the facts that "verifiability" is the wrong word for it and that "not truth" is a redundant and potentially misleading addition.--Kotniski (talk) 11:40, 24 September 2011 (UTC)

I agree, and I take your point that we are now dealing with fallout from the poor choice made in settling on the word "verifiability" way back when.  SlimVirgin said as much, that veritas means truth, but that that is not why the word was chosen.  Meanwhile, we have Jimbo, coming from the viewpoint of a fuzzy thinker, who genuinely wants an accurate encyclopedia, reasonably mis-associating "verifiability" and truth.  IMO, we must strive to simultaneously integrate all of these multiple viewpoints in order to find a common viewpoint.  Unscintillating (talk) 12:53, 24 September 2011 (UTC)

"The flaw in this way of thinking is that, while it may be that editors will nearly always agree on whether a source says something, there is plenty of room for disagreement as to whether a given source (possibly in isolation) is reliable on a given point."

The flaw in your thinking is an obsessiveness about words out of context and an inability to distinguish different issues, under "verifiability, not truth" the way to show a particular soucre is unreliable is by references to other verifiable sources. (other issues of pov and relevance are dealt with elsewhere)

"We do humanity no service if we switch off all our other critical faculties and just report everything we read in any book, journal or newspaper as if it were unquestionably true."

Other critical faculties? We have constraints within this policy and other policies to deal with that. We do humanity no service (is there a more puffed up way you could put it?) by misrepresenting or denying reliable sources because you think you know what truth humanity needs. (eg., Do you have a citation to your service to humanity?)

"IMO it is the premise that the phrase "verifiability, not truth" has a well-understood meaning that is flawed."

If we cannot read in context then there is no point to writing anything at all. Alanscottwalker (talk) 17:27, 24 September 2011 (UTC)

If you can assert that there is no ambiguity in "verifiability, not truth", then you have the unifying knowledge that we need that integrates all of our various viewpoints.  Please lead on and explain what it means.  It should include at least these three viewpoints:

• "verifiability, not truth" is a jolt of psychological force applied against editors with crackpot theories, and means that Wikipedia editors don't want to spend time discussing such theories that are not sourceable.  In this context, editors may refer to truth using scare quotes, "the 'truth' " or as trade-marked, "Truth^TM".
• "verifiability, not truth" means that Wikipedia editors don't want to spend time discussing genuinely useful material that is not sourceable, i.e., being true isn't good enough if we can't blame someone else.
• "verifiability, not truth" means that Wikipedia editors don't want to spend time discussing potentially inaccurate material, i.e., if it can be sourced, then no questions need be asked, expediency is that we publish it and move on.

Unscintillating (talk) 18:05, 24 September 2011 (UTC)

And indeed, even if all established editors had the same understanding of the phrase, that would still be no justification for using it - we are writing for people who don't know yet, not for people who already know, so we must try to use words which really do (in the real world) mean the things we want them to mean, not words which we've got together and decided will mean something special for us. (We do actually define "verifiability not truth" in the same sentence, so my objection to it is more philosophical dishonesty and abuse of language than actual misleading; I think the real ambiguity - which is not explained - stems from the phrase "the threshold".)--Kotniski (talk) 18:27, 24 September 2011 (UTC)

Philosophical honesty and abuse of language. 'There are more things in heaven and earth, than are dreamt of in your philosophy.' Two words have multiple connotations and shades of meaning; they are deduced by context. That's not abuse; that is language.-- Alanscottwalker (talk) 18:40, 24 September 2011 (UTC)

It is if the meanings we give them are so far removed from their normal meanings that they can only be deduced from context if we step back and completely reword what we've just said in words that really do have the meanings we desire.--Kotniski (talk) 19:47, 24 September 2011 (UTC)

So you claim that you understand the meaning of the phrase "verifiability, not truth" using your ability to interpret context, are you then able to explain it?  Unscintillating (talk) 20:53, 24 September 2011 (UTC)

To quote Kotniski: "We do actually define "verifiability not truth" in the same sentence ..." You should read it. Alanscottwalker (talk) 21:05, 24 September 2011 (UTC)

So the problem of ambiguity in "verifiability, not truth" is not a real world problem, but is a problem of the perception of editors that lack skill in interpreting meaning within context, and who would benefit from reading to the end of the sentence.  Did I get that right?  Unscintillating (talk) 21:35, 24 September 2011 (UTC)

The problem of ambiguity arises with almost all words; words have multiple meanings, connotations, and shades of meaning. The approach to claim of ambiguity, is straight forward. The word asserted to be ambiguous cannot be judged in the abstract; it cannot be read alone. It can only be read in context. So, the claim that "verifiability, not truth" is ambiguous fails when it is read in a sentence defining what it means. You may not like the definition, but that's a different issue. Alanscottwalker (talk) 13:11, 25 September 2011 (UTC)

The slogan "verifiability, not truth" is contradictory and nonsensical. It's used out of the context of WP:V. Note that one could similarly have a context which defines the words in the contradictory nonsensical slogans of Orwell's fictional work 1984: "war is peace", "freedom is slavery", "ignorance is strength". It's better writing to use the common meanings of words instead of changing them. The two words "not truth" are not needed in the first sentence. The sentence means the same thing without those two words and wouldn't foster the nonsensical slogan "verifiability, not truth" if they were removed. --Bob K31416 (talk) 14:42, 25 September 2011 (UTC)

Do you read your own overheated examples? If you did, you would know how absurd your argument is. The phrase is not, "verifiability is truth," the phrase is, "verifiability, not truth." They are two different words. Alanscottwalker (talk) 15:04, 25 September 2011 (UTC)

I suggest you reread my message and note what the slogans all have in common: They are nonsensical and contradictory. --Bob K31416 (talk) 15:20, 25 September 2011 (UTC)

I reread your message. It is still absurd because you claim unlike things are the same. They are not; just as, verifiability and truth are not the same, as the phrase says -- they mean different things. Alanscottwalker (talk) 15:37, 25 September 2011 (UTC)

In fact, verifiability and truth are a lot closer in meaning than this phrase makes out (something is verifiable if we can ascertain that it is true, so a reasoned assertion that P is true is more or less the same thing as a reasoned assertion that P is verifiable). But what we are contrasting in this sentence is not verifiability vs. truth at all (as we see by reading the second part of it): it's the published claims of reliable sources vs. the personal convictions of editors. This is just one of the reasons why "V not T" is wrong; and that's just one of several things wrong with this sentence (not to mention the title of the policy).--Kotniski (talk) 18:03, 25 September 2011 (UTC)

Verifiability is the ability to verify -- in this sentence, the ability to verify that a reliable source has said something. Truth is not the ability to verify, it has no ability in it - it's an assertion, an ipsa dixit. In the scheme of what editors are doing, it makes sense that we first require them to come to agreement that a reliable source has said something, and agree what that something is (no misrepresentation or mistake because, we know the truth). This is the humility required of us -- we must first take the source on its own terms, even when (especially when) we disagree with it because we know the truth, because in most (all) cases we don't in fact know the truth, and we should in any case assume we and our readers do not, when first approaching the source. Moreover, in practice, and by the dictates of this policy, one editor cannot convince another editor that they have the truth (don't tell me the truth, that's irrelevant, convince me with evidence). Only then can editors proceed to agree on correct representation of the source, contextual reliability, relevance to the topic, POV and other considerations for putting it in an article. Alanscottwalker (talk) 19:30, 25 September 2011 (UTC)

This is all doubtless the case, more or less, but "The threshold for inclusion in Wikipedia is verifiability not truth" doesn't even begin to express it. (Any attempt to glibly summarize all this without mentioning the key concept "reliable sources" simply must be wrong. As must anything that implies that there is a single threshold - necessary/sufficient condition - for inclusion.)--Kotniski (talk) 08:42, 26 September 2011 (UTC)

Stepping back from philosophy: otherwise reliable sources can be mistaken (not merely misread) - they are then unreliable on that point; but we show this by verifying that the consensus of otherwise reliable sources disagrees; in Chopin's case, adding that he was a French citizen by birth, with a note citing the books that say so. Septentrionalis PMAnderson 18:45, 24 September 2011 (UTC)

Quite right. We don't know, who his father truly is, only his mother could know (and she might not), but, don't despair, we can report what reliable sources have said. Alanscottwalker (talk) 19:13, 24 September 2011 (UTC)

That presupposes we know which sources are reliable on this point. That, unfortunately, is something else we don't truly know. (Remember: we didn't have a source that said directly that he was a French citizen from birth.) This whole idea that we reduce a subjective judgement to an objective one using the magic formula "reliable sources" is what I mean by philosophical dishonesty - we're actually replacing one subjective judgement with another, but we pretend otherwise when we describe what we are doing as "verification".--Kotniski (talk) 19:47, 24 September 2011 (UTC)

Question: why doesn't WP:REDFLAG cover the Chopin-type case? --FormerIP (talk) 20:30, 24 September 2011 (UTC)

It probably should, though as it's currently worded it fails to make clear that it would cover that sort of case - and we were talking about the lead of the policy, which ought to be consistent with the policy (the fact that important exceptions are described, somewhere a long way down the page, is no excuse for making an oversimplification in the lead).--Kotniski (talk) 06:15, 25 September 2011 (UTC)

Oh, you want to discuss "reliable sources." That's different talk page. See instructions, at the top of this page.Alanscottwalker (talk) 21:02, 24 September 2011 (UTC)

Is that directed at me? You might want to clarify, but REDFLAG is a provision of WP:V, not WP:RS. --FormerIP (talk) 21:33, 24 September 2011 (UTC)

No. Sorry, I was responding to the one above you, but I did not want to cut in.Alanscottwalker (talk) 21:39, 24 September 2011 (UTC)

Request for comment

Please join the discussion at Wikipedia talk:Manual of Style/Television#Sourcing plot sections.  BIGNOLE  (Contact me) 15:19, 24 September 2011 (UTC)

A flaw in WP:SPS

The following chain is misused too often:

• Post something on your webpage - say, name, birthdate, birthplace, etc., which disagrees with (reliable) sources.
• Change this information on wikipedia, saying "error correction" per personal webpage.
• After a while, this is picked up by the echo. Done.

WP:SPS does not prevent this, only otherwise - if the 5 criteria are met (and they are), the procedure is fine. Materialscientist (talk) 04:16, 25 September 2011 (UTC)

What do you suggest we do about it?--Kotniski (talk) 06:10, 25 September 2011 (UTC)

Discuss and update the policy page, limiting the use of information from the personal web pages which contradicts reliable sources. Materialscientist (talk) 06:14, 25 September 2011 (UTC)

Um... When you say "Post something on your personal webpage", who are you talking about? It matters. The point of WP:SPS is that a self-published source (such as a personal website) is a reliable source for statements about the author of that self-published source, but is not reliable for statements about other people.

So... it is important to ask who the "you" in your scenario is. Let's say that a notable politician is the subject of a Wikipedia article. That politician looks at the article and notes that we have his birthday wrong. So he tries to correct the information. We respond by saying... "Sorry, but the sources say you were born on the date given, and by our rules we have to follow the sources." So he goes to his personal webpage and posts the correct date. His personal webpage is considered a reliable source for information about himself. So, we now have a disagreement among reliable sources, and we must weigh the relative reliability of the sources (see WP:DUE). This is not a "misuse" of our rules... it is an acceptable way for someone to try to correct misinformation about themselves and set the record straight.

Now, let's change the scenario... Joe Blow is working on the article about the politician. Joe Blow thinks that the article has the politician's birthday wrong, and tries to "correct" it. We respond by saying "Sorry, but the sources say that the politician was born on the date given, and we must follow the sources". So, Joe goes off to his personal website and posts what he thinks is the correct date. In this case, we don't care what Joe posts on his website. Joe's personal website is not considered a reliable source for information about the politician (unless Joe is an acknowledged published expert, writing in his field of expertise). If Joe is just some random guy then his opinions carry no weight. So... if Joe comes back and tries to change the article, citing his website... we would appropriately challenge his change. Blueboar (talk) 13:48, 26 September 2011 (UTC)

Allowing for economics in verifiability

This is a genuine concern affecting the academic community at the moment, so I thought it should be considered on Wikipedia also. Because of the cost of running peer-reviews and especially formally publishing material, it has been observed (in the Times Higher Education Supplement, and other sources) that publishers are simply refusing to review material for reasons unconnected with veracity, simply because it is not profitable for them to do so. As a result, some fields have been forced to change their expectations: for example, academics in English were generally expected to write and publish a book, but this is now becoming an unreasonable expectation because publishers will not publish an academic book that is likely to only sell a few copies, regardless of its quality (also from THES). The fact that Wikipedia benefits from the fact-checking services of these publishers without compensating them and while, in fact, competing with them has been another concern.

Do we need to consider different paths to verifiability that are not so affected by economic and political concerns? — Preceding unsigned comment added by Hyphz (talk • contribs) 18:38, 25 September 2011 (UTC)

• Because Wikipedia is written by non-experts, "verifiability" on Wikipedia means--and can only mean--"able to be checked against reliable, published sources". Academic sources are, for many topics, the most reliable ones we have. We rely on academic publishers being willing to publish high-quality research. If they are unwilling to publish material, then we must conclude that it is unreliable.

  Wikipedia does not compete against academic publishers. We aim to provide a neutral summary of human knowledge on a given topic, but a summary written by anonymous non-expert volunteers does not in any way compete with a reliable academic source.

  It is unfortunately true that money has an impact. Before science can be a quest for knowledge, it must first be a quest for funding, and the only way to make a small fortune in academic publishing is to start with a large one. But Wikipedia is not to blame. Even if it weren't for Wikipedia the internet would have had a massive impact on academic publishing.

  The future for academics wishing to publish important research involves university websites rather than heaps of dead trees, and Wikipedia's policies already reflect this truth.—S Marshall T/C 09:26, 26 September 2011 (UTC)

• Re "Because of the cost of running peer-reviews and especially formally publishing material, it has been observed (in the Times Higher Education Supplement, and other sources) that publishers are simply refusing to review material for reasons unconnected with veracity, simply because it is not profitable for them to do so." - Are you referring to page charges made to authors? If so, these charges existed long before there was Wikipedia. Or are you referring to rejection because journal editors don't feel material is worth peer-review because there would not be enough interest in a particular article if it were published? If your point is that journals reject material because the information would be put in Wikipedia, then wouldn't that be a reason for the journals not to publish any material, not just reject some material?

Re "Do we need to consider different paths to verifiability that are not so affected by economic and political concerns?" - I don't see what those paths could be. What do you have in mind? --Bob K31416 (talk) 17:25, 26 September 2011 (UTC)

I question whether we are "competing with" scholarly publishers. It often seems to me that we are their single most effective advertising service. Not only do Wikipedia's editors buy books and journal articles for the purpose of writing articles, we also cheerfully provide convenient links to the publisher's paywalls. WhatamIdoing (talk) 21:14, 26 September 2011 (UTC)

We have a small overlap, where we could be seen as a competition to academic publishers, that is academic encyclopedias. Though Wikipedia is not a academic encyclopedia as such nor possess a comparable reputation or quality control, it nevertheless is increasingly covering their topics as well (sometimes in more detail even) but with a slightly different slant. However the real comptetition in this area probably comes from academic wikis (scholarpedia, citizendum, various scholarly or science wikis) and from (free) online journals, ebooks and electronic lecture notes where publication, editorial processess and (peer) reviews are facilitated by scholars without the involvement of traditional publishers.

But whatever the future will bring there, WP will simply use the best/reasonable reliable the academic/science community has to offer, no matter whether it is published through traditional publisher or other venues.--Kmhkmh (talk) 21:38, 26 September 2011 (UTC)

Just going back to the original question, questions like this come up often on WP:RSN. Basically this question is probably more suitable there. It is clear that different fields have different types of publications, and some are unfortunately pretty poor. In an extreme case it might even be argued that Wikipedia should not cover a certain subject because the sourcing just is not good enough. This seems to normally happen with things like pop culture and hobbies, like maybe dog breeding, and generally there are constant debates about finding the right approach on WP. Then we have other controversial cases where respected people are publishing complex stuff outside the older styles of peer review system, for example on web pages, or using "crowd sourcing" of some type. This is clearly going to happen more and more, and we simply can not give one answer which covers all fields and possibilities. But I think the general rules are like this:-

• Try to identify and learn about what publications exist for a field, and who are the right types of qualified experts whose opinions are the ones that define the field. There should be some which are relatively uncontroversially respected. Sometimes it might just help to see who newspapers cite and talk to when they have a question. (In the end, if there are no uncontroversial publications of experts we have a problem to cover the field at all.)
• If many of the publications are of un-traditional form, for example web-pages, you should try to get evidence that shows that those particular webpages or whatever are cited by and treated seriously by, the above mentioned recognized experts and publications. This is how I understand the spirit of WP:SPS.
• The basic idea is to make sure that our sources "have a verifiable reputation for fact checking".--Andrew Lancaster (talk) 11:54, 27 September 2011 (UTC)

Has Verifiability become the new Truth?

I think in some editors mind's Verifiability is truth and as a result anything presented on the talk page challenging that truth gets labeled as OR regardless of how reliable it is.--BruceGrubb (talk) 18:55, 29 September 2011 (UTC)

Yes, and particularly for some science articles that go into the techinical details this can be a real problem. This is why I wrote up WP:ESCA some time ago. Count Iblis (talk) 00:46, 30 September 2011 (UTC)

Bruce... perhaps you are not aware that OR is allowed to be discussed and explored on talk pages (especially if that discussion is relevant to the process of editing the article)... so it does not matter whether something that is being presented on the talk page gets labeled as OR. WP:NOR applies to what is written in the article, it does not apply to what we discuss on the talk page. This is true for all of our content polices - in fact its why they are called content policies (they apply to the content of articles). Talk pages are governed by our behavioral policies, not our content policies. So, as long as we don't violate a behavioral policy, we can discuss just about anything on a talk page... even things that a content policy says should not be added to an article (as another example, there is nothing wrong with discussing unverifiable information on a talk page if need be... even though that unverifiable information should not be added to the article). Blueboar (talk) 01:36, 30 September 2011 (UTC)