Wikipedia talk:Verifiability/First sentence/V FS Archive 49

Trying to WP:KISS people

We have a TLDR problem with the 'rules', and just starting an embryonic project to try to get the concept across in a sinlge-sentence 'The simplest explanation' thing. Please seethis thread for background stuff.

The idea is to get (hopefully soon) a template-wossname for all the rules pages, so that anyone with a slightly less comprehensive vocabulary can understand stuff really easily. We have a problem where some of our stuff is written so collegiately that it's just hard for some people to get the point. So ..... could we possibly, pretty please, have the simplest explanation back at the top :o) ? Pesky (talk …stalk!) 07:04, 21 June 2011 (UTC)

• In response to that Village Pump thread, I propose (again) that we simplify and shorten the complex grammar and the more obscure usages in this policy. Specific measures that I propose (again) include:-

  1) Switching from the hortative to the imperative;

  2) Switching from the passive to the active; and

  3) Using the word please.

  The practical effect of these changes would be:- This policy requires that all quotations and anything challenged or likely to be challenged be attributed in the form of an inline citation that directly supports the material → Please attribute all quotations and anything challenged or likely to be challenged with an inline citation that directly supports the material. | Anything that requires but lacks a source may be removed, and unsourced contentious material about living persons must be removed immediately. → You may remove anything that needs a source but lacks one. Please Remove unsourced contentious material about living people at once. | etc.

  Subject to considering other editors' objections, for the time being I tentatively support PeskyCommoner's proposed one-sentence summary as well.—S Marshall T/C 07:21, 21 June 2011 (UTC)

Really need some good brains working on a project to make all our rules easily-understandable by all our readers! Everyone creative enough and willing to help out on this is needed :o) Pesky (talk …stalk!) 07:28, 21 June 2011 (UTC)

Whatever we decide on, we have a responsibility to make sure that the 12-year-old who's just been slapped with a template, etc., can actually understand why. So verifiable, in kid-terms, is "People have to be able to check that you didn't just make it up!" Which adults can also understand, and sensible adults will appreciate why this is included as a very simple explanation. Pesky (talk …stalk!) 08:28, 21 June 2011 (UTC)

Changing "must be removed immediately" to "please remove" is not making the text simpler and better understandable, it is removing its teeth. Fram (talk) 08:40, 21 June 2011 (UTC)

Agreed: "please" makes it sound like a choice, which it isn't. ~ Mesoderm (talk) 08:42, 21 June 2011 (UTC)

Of course it's a choice - no-one can be compelled to do anything on Wikipedia.--Kotniski (talk) 07:31, 22 June 2011 (UTC)

Well, the other reason to include basic courtesy in the policy has to do with reactance. If you want someone to behave in a certain way, they're more likely to comply if you ask them to do it than if you tell them to do it. I think it's important to remember that a lot of people who are asked to read this policy are stubborn, or behaving in a difficult way, already.—S Marshall T/C 09:32, 22 June 2011 (UTC)

That is not true for everyone. Some people respond very effectively to "Remove this immediately", but see "Please remove this" as a recommendation that they may comply with, or not, as they please. Most Wikipedians don't see our "no libel" and "no copyvio" rules as impinging on their freedoms (=a necessary precondition for psychological reactance). WhatamIdoing (talk) 21:30, 22 June 2011 (UTC)

The way I understand reactance is that it's about a contrary reaction to strong social pressure. (I used to work with prolific young offenders, which is the context in which I used to encounter that term.) If I've used the wrong word then I've used the wrong word—what I'm trying to get across is that people who've been directed to read WP:V may well be cussed and contrary. Personally, I'd prefer to persuade rather than command in our policies. But I've already axed that "please" from the draft at Fram's request.—S Marshall T/C 23:08, 22 June 2011 (UTC)

• It says "please remove... at once". You want me to omit the "please"? I'd remind you that this policy is addressing volunteers, not employees.—S Marshall T/C 08:43, 21 June 2011 (UTC)
  • I'm with you here - we can be firm without being obnoxious - there's nothing wrong with "Please". S Marshall's point about volunteers, not employees is excellent. Remember that experienced editors are simply more experienced - not superior beings with a right to boss people around and act arrogantly rather than politely! Pesky (talk …stalk!) 05:18, 22 June 2011 (UTC)

• • It also adresses things we do not want in the encyclopedia, under any circumstances. "Please remove at once" means that if people willfully and knowingly let contentious material on BLPs in articles, there is little we can do about it (taking actions against those editors, I mean, we can still remove the content obviously). It should be made clear that knowingly letting BLP violations stand in articles (or reuintroducing it and so on) is not acceptable. Making the policy more strict in this regard ("must" instead of "please") makes it easier to deal with serious violators who try to wikilawyer their way out of it. Fram (talk) 10:21, 21 June 2011 (UTC)
    • Very well, I've removed the "please". The sentence now reads "Remove ... at once." I don't like it—policies are requests made of volunteers, and our conduct is supposed to be collegial, respectful and courteous—but I recognise the consensus is against me here.—S Marshall T/C 10:32, 21 June 2011 (UTC)
      • Thanks. Some policies are not really requests, but obligations: if you want to have your edits on Wikipedia, you should under no circumstances introduce copyright violations and/or unsourced contentious material about BLPs. Many other things are subject to discussion, but some things are basically non-negotiable. Making it clear in our policies that some rules are stricter than others isn't a bad thing. Fram (talk) 12:34, 21 June 2011 (UTC)

The "KISS principle" (to keep it short and simple) is served with the nutshell template at the top. Cambalachero (talk) 12:55, 21 June 2011 (UTC)

Some of our 'nutshells', though, aren't worded as simply and jargon-freely as they could be. We always need to remember that our intial 'target audience' is comprised of people who don't know the rules, therefore don't understand the basics of the rules, don't know the jargon, and are quite possibly overwhelmed by a page full of jargon-loaded text. So it's important that the first introduction is an incredibly easy-to-understand statement. The very people who are most likely to fall foul of the rules and get into a pickle are the ones we need to be most careful to explain things simply to, to begin with. The 12-year-olds - and they do edit! - who need to know what "verifiability" actually means, and what "neutral point of view" actually means, in terms that they themselves would be likely to use. Pesky (talk …stalk!) 13:07, 21 June 2011 (UTC)

The problem with putting this policy in a nutshell, though, is that we don't really know what the main point of the policy is, or why it's separate from WP:NOR.--Kotniski (talk) 07:40, 22 June 2011 (UTC)

• What? The main point is to make you show your working after you've researched an article. Editors ought to know where you got your facts from. I mean, I do perceive a number of problems with WP:V, but it repeats that main point about six times in successive paragraphs, one after the other—how can it not be completely obvious what the main point is?—S Marshall T/C 08:00, 22 June 2011 (UTC)

• So let's replace the opening sentence with something that says more or less that, and see how many milliseconds it remains before being reverted by someone who thinks the point of the policy is not that at all.--Kotniski (talk) 08:29, 22 June 2011 (UTC)

• Realistically, the first sentence is a sacred cow at this point. Any attempt to change it will be reverted and forced to RFC. The RFC will be inconclusive, which means the first sentence will survive in its current form, because Wikipedian behavioural policies require stagnation in the absence of consensus. As I've explained before, the only way to make significant changes to the first sentence is via a slow accretion of stealth edits, each unobjectionable enough to avoid the "OMG someone's changed WP:V without six pages of discussion! Revert revert revert!" reflex.—S Marshall T/C 09:38, 22 June 2011 (UTC)

• Eeh... no edit to that sentence will be stealthy enough to achieve that :( ... But it doesn't seem unreasonable for us to ask the people who think this page ought to remain in its present form and separate from WP:NOR, to say (with a reasonably unified voice) what the scope and principal message of each of the two pages is supposed to be. Can anyone do that? (Past experience suggests not, but you never know.)--Kotniski (talk) 09:47, 22 June 2011 (UTC)

• Personally I think it's best to keep the positive policy and the negative policies separate. V is the only positive core policy (the only one that says what editors must do). NOR and NOT are negative (because they say what editors mustn't do). Of the three, I find V the least problematic; NOR is much harder. In my one, single, abortive experience at FAC I found that users do want you to conduct original research. Whether the problem was with NOR or with FAC or both is a matter I'm still thinking through, although I suspect FAC is the more problematic area. NOT isn't really a coherent policy at all, it's just a shopping list of things most editors think other editors shouldn't do; there's no form or structure or discernible flow of thought to it.—S Marshall T/C 10:56, 22 June 2011 (UTC)

• What is it you think that editors must do?--Kotniski (talk) 12:07, 22 June 2011 (UTC)

• Show their working. To me, a verifiable article is like a scientific experiment: it's not enough to show your results, you have to show your readers how to duplicate them. (Don't parse this analogy too closely. I'm well aware that experiments are very often original research. In Wikipedia article-writing you've got to steer between the scylla of original research and the charybdis of replicating what your sources say too closely. WP:V ought to help you walk this tightrope.)

  It follows that for me, there's more to verifiability than citing sources. To the extent that's appropriate for an informed lay person, you can go beyond what your sources say. For example, if your source says that a moon crater is 100km in diameter, you can say that its area is about 7850km². (You're allowed to do high school maths. It's not original research if Euclid proved it in 300 BC.) And if you're translating a source, you can elaborate on it to give context. (The German words ertrinken and ertränken have different meanings, but they both translate into English as "drown". A simple translation of a news piece might make a tragic accident sound like a premeditated murder, or vice versa. Du and Sie both translate as "you", while fressen and essen both translate as "eat", so a simple translation of "du frißt" can make a quite contemptuous remark sound like an unexceptionable statement of fact.)

  I'm in danger of going a bit far afield. The answer to your question is that Verifiability is a behavioural rule that seeks to require editors to behave in a certain way.—S Marshall T/C 12:56, 22 June 2011 (UTC)

• Yes, I see what you're getting at... except that editors aren't really required to do anything. You only need to "show your working" if someone asks for it, and even then, you won't be punished if you don't (although what you wrote might be removed). I suspect what it comes down in effect to is that you mustn't keep re-adding material that you can't source - but now we're back to a negative again.--Kotniski (talk) 13:56, 22 June 2011 (UTC)

• Removing someone's writing is a punishment, though. I know what the wikidealism is, but the down-and-dirty reality is that when someone cares about something, they'll fight to defend it, and they may become angry or humiliated if it's taken away from them. And the "something" can be a word or a phrase in an article. (Or a policy.) And there's this business of anything "likely to be challenged". In fact, you show me anyone who's got any real experience of Wikipedia at all, and I'll show you someone who automatically cites the best sources they can when writing an article. It's a behaviour reinforced by experience of seeing people's writing removed.—S Marshall T/C 14:10, 22 June 2011 (UTC)

Kotniski, you're kinda dead-right about the WP:V = WP:NOR thing. It's two sides of the same coin - which became abundantly clear to me when I was trying to think of how to put NOR into as simple terms. But ... (yes, how horrid, I started a sentence with a hanging "but", lol!) people tend to get directed to these two pages from slightly different behaviours and levels, I think. The V thing covers "you can't just make it up, or look as though you've made it up", and the "OR" thing is more "even if you're right, even if you're an expert, it has to be published elsewhere and / or by someone else (RS, too)before we can include it". Scholarly researchers know the importance of verifiability; they and their research teams do, and explain, their own verification in their scholarly papers, and the WP:V page is more likely to irritate them ("I know that, I've explained the verification!") than the WP:NOR page .... ("Ah! OK, I can't publish my paper in here!") Pesky (talk …stalk!) 03:58, 23 June 2011 (UTC)

I remain unconvinced; my experience is (if anything) that it's the other way round - if people look as if they're making stuff up, they usually get directed to WP:NOR. They get directed to WP:V usually if there's a specific problem with sources. In fact you can get directed to either policy in either situation (that's assuming there is any real difference between the two situations) - the two pages are forks of each other, pure and simple, with information either duplicated or divided randomly between them.--Kotniski (talk) 10:10, 23 June 2011 (UTC)

• I genuinely don't think they are, Kotniski. WP:V is the principle that each individual fact must be sourceable. WP:NOR is the principle that each individual conclusion drawn from the facts must be sourceable. WP:V is about what you should do, and WP:NOR is about what you shouldn't do. They really do seem like separable matters to me.—S Marshall T/C 10:19, 23 June 2011 (UTC)

I really don't see any difference. Structurally they both say that every statement must be sourcable; I think that they 90% overlap with each other. The 10% unique to wp:nor is explaining/reinforcing that WP articles aren't the place for creativity, research, presenting new ideas, drawing conclusions etc.. It also coined/coins the noun ("OR") for most violations of wp:ver. North8000 (talk) 11:13, 23 June 2011 (UTC)

That makes no sense to me. How is "each fact much be sourceable" something you should do, while "each conclusion...must be sourceable" is something you shouldn't do?? It seems everyone who thinks V and NOR should be separate has their own private vision of what the separation ought to be; but these visions are never consistent either with each other, or with the actual arrangement of information we see on the two pages, or even with the meanings of the phrases "verifiability" and "original research". I see it the same as North8000.--Kotniski (talk) 14:25, 23 June 2011 (UTC)

Because V is "follow the sources" (positive) and NOR is "don't go beyond the sources" (negative).—S Marshall T/C 18:34, 23 June 2011 (UTC)

I don't see the difference. It's the same statement, phrased either using a positive or using a negative. It doesn't seem to justify having two separate pages.--Kotniski (talk) 19:19, 23 June 2011 (UTC)

I just deleted the "So" from the beginning of the second sentence of the nutshell. So does anyone have an objection to that? ;-D --Tryptofish (talk) 20:04, 22 June 2011 (UTC)

That's fine! I put "this means that" at the beginnning of it, instead, just to kinda make the link :o) Pesky (talk …stalk!) 03:46, 23 June 2011 (UTC)

• Some years back SlimVirgin (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) noticed the conceptual overlap between V and NOR, and came up with a proposal for merging the two policies at WP:ATT. I wasn't around when this was a serious proposal, but apparently it got a lot of support before it was shot down by Jimbo. For my part, I think it's an excellent idea and a good way to simplify the policies without diluting them substantively. Whether we decide it's a good idea for us to resurrect the proposal (FWIW I would throw my support behind any efforts to do so), the history of it may inform our decisions about what to do here. --causa sui (talk) 18:38, 23 June 2011 (UTC)

I agree. Except I would say that NOR really needs to be chopped up and distributed into both WP:RS and WP:V rather than creating a single new policy WP:ATT. For instance, WP:PRIMARY, WP:SECONDARY, WP:TERTIARY really belong in WP:RS, because they each talk about how certain qualities of sources affects their reliability. On the other hand WP:SYNTH belongs as a subsection of WP:V, because it tells us how we can and can't use sources, once we've determined that they are reliable. But, NOR really seems to serve no purpose other than stating WP:V in a negative sense, and covering WP:SYNTH. ~ Mesoderm (talk) 19:25, 23 June 2011 (UTC)

Guidelines re verifiability of the content of reported statements?

Are there guidelines (either for articles in general or for information about living persons) that address the situation:

1. A reliable published source says that So-and-so said Such-and-such.
2. No reliable source directly says Such-and-such, so Such-and-such is unverifiable and may even be false.

For example, a Wikipedia article says that So-and-so said Such-and-such, accompanied by an inline citation to a reliable published source. The material is challenged on the basis that Such-and-such might not be true, even though So-and-so said it. The challenge is rebuffed on the basis that Wikipedia is not saying Such-and-such, it is merely saying that So-and-so said Such-and-such, which is verifiable and not in dispute.

Here's a made-up scenario. A notable riot occurred and a reliable source quotes Paul Prominent as saying "Steven Stirrer incited the riot". There is no doubt that Paul made the statement about Steven. However no reliable source says that Steven actually incited the riot – and he may or may not have done so. If Steven did, no real harm is done by Wikipedia reporting that Paul said what he said. If Steven did not incite the riot (which we are not sure about), it may be considered a slur, no matter that Paul genuinely said it.

Are there guidelines that cover this? It would be easy to use a genuine quotation (or paraphrase of a quotation) to slip in material that would not stand if made outside a quotation (or paraphrase of a quotation). Thanks. Nurg (talk) 06:40, 26 June 2011 (UTC)

It... depends. Context makes a huge difference in the sort of scenario that you're outlining above. If Paul Prominent is the cities' mayor (and it's at least a moderately large city), or some even more important public official, then including such a statement is probably justified regardless. Otherwise, you're getting into the realm where considerations of due weight start coming into play.
— V = IR (Talk • Contribs) 05:27, 28 June 2011 (UTC)

An encyclopedia is mainly for truth, not for sources. Search machines are for sources. Truth has more prerequisites than sources. Sources, just as witnesses, are accountable or not. Sources have biases, different readings. Sources may not be just written text but any human sensed, in-net existing media, supporting the case. Therefore, in this specific case, nothing at all should be written on Steven. Nothing as long as there is no verdict, meaning a much reliable source, a taped telephone call of Steven for example, or many differently originating sources, or the opinion of many. But we can write on Paul's statements, in an article about Paul, nowhere else. That's my opinion. Alkis0 (talk) 18:52, 30 June 2011 (UTC)

Re Wikipedia and truth, see the initial paragraph of WP:V. Re Wikipidia on Steven and what Paul is reported by reliable sources to have said about him, see WP:BLP. Wtmitchell (talk) (earlier Boracay Bill) 21:55, 1 July 2011 (UTC)

In addition to the obvious (verifiability is a requirement) Do you read that first sentence as including a statement that accuracy should not be an objective of editing? North8000 (talk) 13:37, 5 July 2011 (UTC)

Ohms has it right... context is important. Because we can verify that Paul made a statement about Steven, we have passed the first test: Verifiability (with the caveat that we have not verified that Steven actually did what Paul said... we have only verified that Paul said he did). However, there are still several other tests we must pass before we include this information... the most relevant being Due Weight (and, potentially WP:BLP). A lot depends on who Paul is/was. It also matters which article we wish to add the information to (ie are we thinking about adding the information to the article about Paul, the article about Steven, or an article about the riots? We would give Paul's comment different weight in each of these articles... so it might be appropriate to discuss it in one article, and omit it from another).

In other words... the fact that something is verifiable does not mean we must include it... it simply means that the potential to include it exists. If the something is not verifiable then we do not have that potential. Blueboar (talk) 14:31, 5 July 2011 (UTC)

We ought to explain that somewhere. Few understand that, and of those that do, 1/2 "forget" it when it serves their purpose of the moment. How many times have you seen "reverted removal of sourced material" which in essence says that it is improper to remove sourced material which means that if it sourced, that is sufficient grounds to force it to be included. North8000 (talk) 15:57, 5 July 2011 (UTC)

Yup. This is the problem - when there's some bit that is getting challenged on WP:V grounds, people forget the other hundred and a half pages of policy and guidelines we have on article content. Verifiability is necessary, not sufficient. Regards, causa sui (talk) 16:08, 5 July 2011 (UTC)

Plus the unwritten one which gets immediately thrown under the bus whenever there is a dispute: Editors strive to write quality, informative accurate material. North8000 (talk) 21:15, 5 July 2011 (UTC)

Polls: First sentence

Poll: Misleading opening statement

• For whatever reason, this statement, "The threshold for inclusion in Wikipedia is verifiability, not truth: whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true." is problematic and needs to be rewritten:

Support

1. . Casliber (talk · contribs) 12:52, 18 April 2011 (UTC)
2. For too long, this wording has been used to justify the deliberate inclusion of information known to be incorrect. It needs rethinking. User:Short Brigade Harvester Boris (talk) 14:06, 18 April 2011 (UTC)
3. Yes, there are multiple problems with it, as noted in previous discussions - it's only acceptable if you happen to know what it's trying to say, and it is intended (obviously) to be read and understood by people who don't know beforehand what it's trying to say.--Kotniski (talk) 14:36, 18 April 2011 (UTC)
4. The only reason for keeping this misleading sentence would be if there were no good alternatives. However, it is very easy to think of alternative formulations that do an even better job of making clear that we're after the truth as can be distilled from reliable sources, here on Wikipedia. Count Iblis (talk) 14:50, 18 April 2011 (UTC)
5. Verifiability requirements can be strongly stated without double-dissing the concept of accuracy. The first diss is using the straw-man problematic word "truth" instead of "accuracy" and second by inserting the "not" statement in the lead sentence. North8000 (talk) 15:06, 18 April 2011 (UTC) (moved from below Unscintillating's comment) The lead states with emphasis that what we want is "not truth", and so that is what we are getting. (Unscintillating said it well) Time for a change! North8000 (talk) 00:28, 25 April 2011 (UTC)
6. Weak support as I am not sure that rewriting it is the only way of solving the problem. Actually we have two problems: (1) Editors who really and honestly believe that we should make Wikipedia say something that we know is not true. Just because reliable sources agree it is true and we want to be consistent. (2) Editors who pretend to be of type (1) when it fits their agenda. It saves them from agreeing with a consensus that they cannot plausibly disagree with.
   Both problems are relatively rare but should be addressed. I don't care whether this is done by changing the text or by adding a clear explanation that (1) is not the intended meaning. Maybe neither is needed, but just a strong consensus in this discussion, to which we can then point whenever the matter comes up again. Hans Adler 16:21, 18 April 2011 (UTC)
7. Support. "Verifiability" to this point hasn't even been defined in terms of whether a reliable source has actually been cited, or whether a reliable source could easily be found and cited ("Abraham Lincoln was an important figure in the American Civil War."). Moreover, the world "truth" in the phrase not truth has been perverted: it apparently refers to NOT an editor's idea of truth if it cannot (even in theory) be supported by a reliable source. Which is an extremely odd use of the word "truth," and a very bad way to use it. The concept invoked is something like "a personal controversial version of truth in the WP editors' mind, that could not be supported with a reliable source." THAT is what WP deprecates, but calling that thing "truth" is an abomination, and an insult to truth. WP does seek truth (what good is an encyclopedia that does not?) It just doesn't seek "personal truth." Editors are asked to keep that to themselves.SBHarris 18:58, 18 April 2011 (UTC)
8. Suggest shortening to "The threshold for inclusion in Wikipedia is verifiability, not truth: whether readers can check that material in Wikipedia has already been published by a reliable source," dropping the words "not whether editors think it is true", because I've seen them misused to dismiss demonstrably well-founded concerns about source accuracy. See #Proposal 2, below. --JN466 16:21, 19 April 2011 (UTC)
9. The lead states with emphasis that what we want is "not truth".  This is what we are getting in the encyclopedia, "not truth".  Unscintillating (talk) 00:01, 25 April 2011 (UTC)
10. Support - The problem isn't with truth being in Wikipedia, everyone wants that. The problem is with what some editors think is true, which may in fact be FALSE. The phrase "verifiability, not truth" is misleading. Remove the "not truth" part. The phrase "not whether editors think it is true" at the end of the sentence is correct and right on the mark. Also, "The threshold" is ambiguous and may mean it's enough to just to be verifiable in order to be included in Wikipedia, which is definitely not correct and everyone here agrees that verifiability alone is not enough to be included in Wikipedia. There's NPOV, etc. This can be fixed by changing "The threshold" to "A requirement" or "A minimum requirement". Please see Proposal 4 below. 75.47.143.156 (talk) 14:13, 27 April 2011 (UTC)
11. Support—Truth is too subjective anyways, and has been used by the fringers to their benefit. OrangeMarlin ^{Talk• Contributions} 00:02, 2 May 2011 (UTC)
12. The word "threshold" is problematic, at least.—S Marshall T/C 23:07, 30 May 2011 (UTC)
13. For what it's worth. The Blade of the Northern Lights (話して下さい) 01:52, 18 June 2011 (UTC)
14. Support. For years I have wondered why it seemed to be saying that it didn't matter if something was true or false. I was sure it didn't mean that, but it is a clumsy way to say what I think it means. This is a core policy that that new editors should read, and they might not read far down the page, so the first sentences are critical. To be told "not truth" is confusing to me, let alone new editors. "Not truth" can be explained (in rather more words) further down the page. Nothing puzzles me more about Wikipedia than that this wording has persisted so long. Bizarre, frankly. Opponents of changing it should put themselves in the shoes of the average novice editor and look at it with fresh eyes. Nurg (talk) 00:40, 25 June 2011 (UTC)
15. Support: The current wording is misleading, and might indicate to some people that we don't care whether or not we are including untrue information, or that we aren't striving for factual accuracy. What does the "not truth" add, really? We are trying to tell people that even if they "know" something to be true, they still must cite a reliable source. However, the instruction "Always cite a reliable source.", is inclusive of the situation where you know something to be true. It says that you always must cite a source. Therefore any statements that say "Even when ..." are extraneous. Why don't we just get rid of "not truth" altogether, and tell people that everything has to have a source, period? If someone tries to add something without a source, and says "But it's true!", then we can say "Doesn't matter. WP:V says you must always have a source." The "not truth" part is unnecessary. ~ Mesoderm (talk) 06:50, 28 June 2011 (UTC)
16. Support: Verifiability is a means for Truth and an encyclopedia is mainly for truth, not for sources. Search machines are mainly for sources. Alkis0 (talk • contribs) 18:22, 30 June 2011 (UTC)

Oppose

1. . It's neccesary to mention that debates for inclusion don't depend on whether something is true or not. Truth is highly subjective, and endlessly arguable. Verifiabilty can be easily checked. If we imply that truth is a matter of consideration in our decision making process, we will encourage original research, endless arguments, and walls of text. We'll never reach consensus on anything. LK (talk) 13:09, 18 April 2011 (UTC)
2. I fear that changing this wording opens the door to unwarranted promotion of fringe theories which is still a major problem here at Wikipedia. In fact, the latest Quarterly Newsletter of the Association for Skeptical Enquiry[1] discusses the problem and actually recommends people stay away from Wikipedia because of the difficulty in dealing with fringe theories. Let's face it. There's a good reason why we don't care about The Truth©: people can argue endlessly over what's true but checking to see if a source says something is a much easier debate to settle. A Quest For Knowledge (talk) 13:48, 18 April 2011 (UTC)
3. I don't think the statement itself is problematic, and I don't think it should be removed or changed... However, I think the explanation of it may be incomplete. As written, it correctly excludes unverifiable information, even if it is "true". What it is missing is a follow up statement on what to do about clearly untrue (or inaccurate) information that happens to be verifiable. Blueboar (talk) 13:48, 18 April 2011 (UTC)
4. Nothing's broken as far as I can see. --Nuujinn (talk) 16:07, 18 April 2011 (UTC)
5. I expect to need the words "threshold" + "verifiability, not truth" in the foreseeable future. My evolving intensity of preference is informed by lessons learned the hard way. --Tenmei (talk) 17:36, 18 April 2011 (UTC)
6. There's nothing wrong with the current wording, and changing it will open the floodgates to every crank who thinks they know the TRUTH™. Even now we are inundated with them, but this wording at least helps mitigate the worst of it. Jayjg ^(talk) 22:54, 18 April 2011 (UTC)
7. The idea that "verifiability, not truth" is the threshold for inclusion is widely used and well-understood on Wikipedia. Some people here are saying there have been attempts to insert material known to be false because of it, but I've personally never seen an example of that in over six years of regular editing; and if such examples do exist, they are rare. For the most part, the idea makes clear to editors that what we do on Wikipedia is supply a survey of the relevant literature, regardless of our personal views. That's not just a means to an end (where what we're really doing is aiming for "truth"), as others have argued. Offering a good summary of the appropriate literature is an end-in-itself. SlimVirgin ^{TALK|CONTRIBS} 01:24, 19 April 2011 (UTC)
8. Wording is fine. Like I've said before, Wikipedia's policies don't currently allow individual editors to assert personal authority over what is true or not. We're only allowed to declare something as true if it says so in a reliable, verifiable source. Therefore, verifiability trumps whatever we personally feel to be true. Cla68 (talk) 02:08, 19 April 2011 (UTC)
9. It's fine, and we understand what it means. (Those who don't can be pointed at Wikipedia:Verifiability, not truth.) And if you need to see the problems with demanding that articles present "the Truth™", then I recommend that you spend a while hanging out at articles about mental illness, where people occasionally name "personal experience" as a "citation" for claims about (for example) the laws for involuntary commitment in their home countries. There's an ongoing dispute in articles related to saturated fat about whether the mainstream view (eating a lot of saturated fat is bad for the heart) has been completely wrong for decades. WhatamIdoing (talk) 02:31, 25 April 2011 (UTC)
10. The statement is fine because "the truth" can only be proven via verifiable reliable sources. Anyone can go and claim that something is not "true" and remove it from an article even if it's well sourced, that's why wikipedia is not about truth. --Enric Naval (talk) 13:37, 27 April 2011 (UTC)
11. As has been said above, surely better than I can say it, the present wording is fine. As "truth" so often depends on the viewpoint of the speaker, we have to use the standard of whether or not something can be verified from a "reliable source", and "threshold" is a succinct way of saying that verifiability is a condition that must be met for inclusion in Wikipedia, but doesn't guarantee inclusion. -- Donald Albury 09:34, 4 May 2011 (UTC)
12. Per SlimVirgin, Blueboar, et al. --causa sui (talk) 19:33, 16 June 2011 (UTC)
13. IT is core to the encyclopedia. As the community grows, more people will come not understanding this core principle, and not encountering lots of people to explain it to them. Misunderstanding core policy is an inevitable consequence of growth. The solution is not to change the policy that has contributed so much to WP's success and credibility, but to explain it better and to promote it more consistently. Slrubenstein | Talk 01:08, 17 June 2011 (UTC)
14. Oppose per the above. And if it is used to "justify the deliberate inclusion of information known to be incorrect," then that would need to be addressed elsewhere (and I agree that it should be addressed, if not judged as outright vandalism!). This part of policy is very clear and needs to be strongly stated as it now is. Dreadstar ☥ 03:13, 17 June 2011 (UTC)
15. Oppose. It says exactly what it means. It has the added value that it is shocking, and makes the reader start to think. And it has years of use and tradition. BE——Critical_{__Talk} 08:04, 17 June 2011 (UTC)
16.  — V = IR (Talk • Contribs) 05:07, 28 June 2011 (UTC)

Discussion

• As above, my beef is that it creates an artificial dichotomy of truth and verifiability as distinct endpoints (which they are), but what needs to be emphasised is verifiability is a means to an end. Casliber (talk · contribs) 12:52, 18 April 2011 (UTC)

Comment: Amplifying Casliber's opinion, please consider these factors. --Tenmei (talk) 17:56, 18 April 2011 (UTC)

• WP:V "frames" each of our edits by providing generally practical, flexible templates which were crafted to help me and others find ways (a) to escape conflict and (b) to move towards areas of identified agreement.
• WP:V communicates in a way that logic alone cannot, especially when people are talking past each other at cross purposes and when a mismatch develops like a "chicken talking to a duck" (鸡同鸭讲 or 雞同鴨講) or when a white horse is not a horse (白馬非馬).
• WP:V is most needed because " ... the simplest thing cannot be made clear to the most intelligent man if he is firmly persuaded that he knows already".

• Some people above seem to be under the impression that "verifiability" (or "whether a source says something") is an objective matter. It isn't, of course - determining whether a source is "reliable" in a given instance is no less a subjective process than determining whether a given statement is "true" (in fact it quite often comes down to the same thing - we conclude that a source is unreliable if the statements it's making appear not to be true). And pushers of fringe theories can exploit verifiability too - by insisting that the sources that support their viewpoints are just as reliable as those that oppose them (or even making WP reproduce claims from fringe sources as the truth, just because no-one happens to have found a mainstream source that specifically contradicts the claims in question).--Kotniski (talk) 14:46, 18 April 2011 (UTC)

Comment: Responding to the above and answering Kotniski's question here: Yes, we can not improve the wording of a significant sentence by deleting the key words "threshold" + "verifiability, not truth". The word "threshold" implies movement and the beginning of a process. This conceptual "threshold" emphasizes the pivotal distinction between (a) a fact which supported by WP:V + WP:RS and (b) a mere factoid which is associated with zero cited confirming support. Adopting Kotniski's words from an archived thread: yes, "in actual fact we do care about the truth of statements and don't mindlessly copy apparent errors from sources"; but this concern only addresses one of a series of plausible follow-up questions. This survey is about averting consequences which attend throwing out the baby with the bath water. --Tenmei (talk) 17:03, 18 April 2011 (UTC)

Couldn't "threshold" just as likely imply the end of a process? Passing a literal threshold means you've entered the house - you're home, dry, and can finally relax in front of the snooker. (And of course something doesn't become a "fact" by virtue of being supported by "reliable sources", or a "factoid" by not being so supported - I don't really know what you're trying to say with that.)--Kotniski (talk) 17:48, 18 April 2011 (UTC)

Q.E.D. -- compare what Blueboar wrote here. --Tenmei (talk) 19:08, 18 April 2011 (UTC)

• What about "Verifiability: The only practical way to approximate the objective truth"? Count Iblis (talk) 14:59, 18 April 2011 (UTC)

• Poll Results: After a week, it seems clear that the poll on this proposal is coming down to "no consensus", with roughly equal support and oppose views expressed. This usually means we default to "Keep as is". Do we need to continue, or shall we accept that the proposal is not going to be adopted? Blueboar (talk) 14:36, 27 April 2011 (UTC)
  • I'd be happy if people could see that they've lost when they plainly have... but I know that my own optimism and faith in my fellow editors sometimes prevents me from seeing such things myself when I'm on the other side. Consequently, I think we can reasonably expect another week of time-wasting arguments here. WhatamIdoing (talk) 03:39, 29 April 2011 (UTC)
    • Indeed. As previous proposals have been retrieved from the archives and restored to the page, I'm restoring this poll too. SlimVirgin ^{TALK|CONTRIBS} 11:33, 20 May 2011 (UTC)

We go with the source
of course of course
and no one may judge when the source is horse
that is of course unless the source is just inside your head
So go straight to the source and if it's horse
that's simply the info you must endorse
we're always on a steady course
repeating what's been said...
screwing with your head.
BE——Critical_{__Talk} 18:16, 17 June 2011 (UTC)

Proposal for a change in the first sentence

Shall we remove the words "not truth" from the first sentence of Wikipedia:Verifiability? 20:32, 10 June 2011 (UTC)

I've been gone and it got archived under the 5 day setting, but below is the lead proposal that emerged from the process here over the last few months. So now it is time to propose the change, with the alternative being "no change" We discussed the desirability of casting a wider net for input (which also means a longer time period for input) Sincerely, North8000 (talk) 14:01, 10 June 2011 (UTC)

Proposal

Replace the entire first (one sentence) paragraph of wp:ver with:

The threshold for inclusion in Wikipedia is verifiability; that is, whether readers can check that material in Wikipedia has already been published by a reliable source. No other consideration, such as assertions of truth, is a substitute for verifiability.¹

...

Notes

1.^ For continuity, the previous version of this text read, "The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true."

The footnote would go in the notes section at the end of the policy and remain longer term.

Support

1. Support This change strengthens wp:verifiability by providing a clearer statement. At same time it corrects the problem that the current wording disparages the concept of striving for accuracy, and the negative impacts that such has had. The disparagement is indirect here, but much worse when that portion is taken out of context and used as a chant, as it often is. Sincerely, North8000 (talk) 14:01, 10 June 2011 (UTC)
2. Support I think it's much clearer than the the current version. Laurent (talk) 14:08, 10 June 2011 (UTC)
3. Support as a move in the right direction (though there are still things wrong with it, as I have pointed out many times).--Kotniski (talk) 14:14, 10 June 2011 (UTC)
4. Support. My position is that this change is necessary but not sufficient, and I would prefer to see the word "threshold" changed as well.—S Marshall T/C 20:56, 10 June 2011 (UTC)
5. Support - an improvement - as I have said on this page before, the construction of a false dichotomy between "truth" and "verifiability" is misleading. Casliber (talk · contribs) 22:52, 10 June 2011 (UTC)
6. Support; many people here seem to fundamentally misunderstand what "verifiability, not truth" means, and it causes far more problems than it should. The proposed wording will eliminate that problem, and it won't enable the "truth warriors" because it says the exact same thing (that information must be verifiable) without allowing for the opposite extreme (people with a severe case of literal thinking who believe "verifiability" means we must source everything to secondary sources, when sometimes a primary source is more reliable, and that we have no respect for the truth and blindly follow the sources even when they're obviously wrong) to impose their standards. The Blade of the Northern Lights (話して下さい) 23:35, 10 June 2011 (UTC)
7. Support. In general, we do care that our articles are accurate, not just verifiable. The "not truth" language causes more confusion than it solves. — Carl (CBM · talk) 02:15, 11 June 2011 (UTC)
8. Support  (1) The lede states with emphasis that what we want in Wikipedia is "not truth".  It appears that some editors really believe that "not truth" is acceptable for the content of articles.  (2) The current text is not useful to explain to a reasonable editor why he/she can't just fix an article to say what a consensus of editors agree is true.  These are both unintended consequences of using a figure of speech in technical writing.  The solution is to remove "not truth" from what verifiability is about, and let WP:Editing policy bear the weight where it says, "on Wikipedia a lack of information is better than misleading or false information".  Unscintillating (talk) 00:59, 12 June 2011 (UTC)
9. Support. It is my opinion that this is wording has been responsible for much of the argumentation and downright silliness on en:wp. John lilburne (talk) 16:32, 12 June 2011 (UTC)
10. support. heaven forbid wp be accused of truth. Darkstar1st (talk) 16:52, 12 June 2011 (UTC)
11. Support This version says the same thing as the prior did, just in a more explanatory fashion than the often confused and misused "Verifiability, not truth" statement. Might I also note that, looking through the first five archives of this page, I saw no consensus or discussion at all about the addition of the statement or the section, which was added by Slimvirgin back in 2005. Silverseren^C 19:42, 12 June 2011 (UTC)
12. Support It's an improvement. The place for iconic crap is T-shirts and tourist traps. SBHarris 22:02, 12 June 2011 (UTC)
13. Support. I see more and more established editors arguing that striving for truth is not necessary, all we have to do is make sure someone else has said it before. That attitude fosters poor editorial judgment, given how often newspapers (and other sources) get things wrong. It's a vital change for this project. See WP:OTTO. --JN466 22:39, 12 June 2011 (UTC)
14. Support. The addendum "not truth" is redundant, potentially misleading, and does appear to have misled some editors into holding the absurd notion that an untrue statement can be "verifiable". By the standard definition given in any decent dictionary, verifiability is a stronger notion than truth and automatically implies it—it's simply impossible to verify something that isn't true. Strictly speaking, what you verify when you check a source which is cited as justifying an assertion of a fact X is not the fact X itself, but the fact that the source asserts X. If many reliable sources assert X and none contradict it, or even if a single highly authoritative source asserts X, and there is no other reason to suspect that X is false, it's reasonable to take X as having been provisionally verified. But if further evidence (which, for Wikipedia's purposes has to take the form of citable, reliable sources) were to be found to cast doubt on the truth of X it would then be no longer justifiable to claim that X itself is verifiable merely because there are some reliable sources which assert it. David Wilson (talk · cont) 17:56, 13 June 2011 (UTC)
15. Support, per most of the arguments above, but I do think that voting on such proposals is not enough to change things; what is needed is to come up with a new draft for the entire policy page, perhaps also NOR and the other core policies, and put those to a vote. Count Iblis (talk) 18:26, 13 June 2011 (UTC)
16. Support. Either remove it or change it to something along the lines of "Verifability, just as much as truth". Like Casliber points out, the current wording suggests a false dichotomy, where both verifiability and truth together should be the basis for the inclusion of any assertion. The current wording is the second most frequently abused formulations in all P&G, right after the fictitious "​neutral" point of view. --213.196.218.59 (talk) 14:45, 15 June 2011 (UTC)
17. Support. I don't much care whether it is perfect. It is a lot better. Nurg (talk) 00:40, 25 June 2011 (UTC)
18. Support It's clearer, and so more immediately-understandable; it's not perfect, but it's definitely an improvement. Nothing stopping us progressing towards still more improvement, and I think a first step towards perfect is always a good idea. Pesky (talk …stalk!) 04:05, 5 July 2011 (UTC)

Oppose

1. Oppose. There is already a straw poll at the top of this page that gained no consensus for change. This may seem like a small tweak, but it's a significant change in emphasis from the current first sentence. The current version has had consensus for years and reads: "The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true." The phrase "Verifiability, not truth" is iconic as a representation of Wikipedia's sourcing and neutrality standards, and there would have to be strong and wide consensus to change it. SlimVirgin ^{TALK|CONTRIBS} 15:41, 10 June 2011 (UTC)
2. Oppose -- as things stand Wikipedia gets warriors who are sure they know what the truth is; WP:V is an essential tool for ensuring that articles are written in ways that reflect sources rather than editors' beliefs about truths. As SV says, the proposed change seems small but is enormously consequential and should not be adopted lightly; it might seem like a way to solve some problems, but unintended consequences loom large here. Nomoskedasticity (talk) 20:42, 10 June 2011 (UTC)
3. Oppose, this needs to stand as-is, as Slim and Nomoskedasticity point out, this is a very consequential change for the policy that will do nothing but enable the truth-warriors. Verifiability, not truth is an important touchstone for the inclusion of material in Wikipedia and shouldn't be diminished by this major change in wording. Dreadstar ☥ 21:02, 10 June 2011 (UTC)
4. Oppose. I feel very strongly that the phrase "verifiability, not truth" is one of the most powerful ideas behind the success of Wikipedia. It's counter-intuitive to new editors, and the ability to explain it to them—as clearly as possible—has been invaluable countless times in avoiding needless disputes. --Tryptofish (talk) 22:05, 10 June 2011 (UTC)
5. Oppose. It's not broke, so don't fix it. A Quest For Knowledge (talk) 22:26, 10 June 2011 (UTC)
6. Oppose, "Verifiability, not truth" is a good touchstone, I see no gain from the change proposed. --Nuujinn (talk) 22:36, 10 June 2011 (UTC)
7. Oppose because the principle that "just because it's true doesn't make it fit for Wikipedia" is rather important and oft-quoted. ╟─TreasuryTag►Regent─╢ 22:40, 10 June 2011 (UTC)
8. Oppose—the policy has long been "verifiability, not truth" and that's an important distinction that's been enshrined in our culture around here. — Preceding unsigned comment added by Imzadi1979 (talk • contribs) 23:02, 10 June 2011 (UTC)
9. Oppose - Prefer the current language. This is a vital phrase in combating POV warriors and fringe theory pushers. Blueboar (talk) 00:18, 11 June 2011 (UTC)
10. Oppose Not broke, don't fix. AndyTheGrump (talk) 22:22, 11 June 2011 (UTC)
11. Oppose nothing wrong with the current wording. --Six words (talk) 23:25, 11 June 2011 (UTC)
12. Oppose for reasons informed by lessons learned the hard way --Tenmei (talk) 01:28, 12 June 2011 (UTC)
13. Oppose - the phrase has been a policy since 2005 (originally in WP:NOR).[2] I don't see a strong enough reason to delete this very stable part of a core policy.   Will Beback  talk  03:00, 12 June 2011 (UTC)
14. Oppose. Current wording is an essential part of policy, removing it will open the floodgates to endless wars among different individuals each possessing their own personal TRUTH™. Jayjg ^(talk) 04:25, 12 June 2011 (UTC)
15. Oppose - I think the current statement is essential to WP. -- Donald Albury 20:49, 12 June 2011 (UTC)
16. Oppose the present wording has done us very good service, and built up a large body of explanations and interpretation, which we should not lose. (Like others here, I refer to Slimvirgin's statement in the poll in the prior section as the best explanation) DGG ( talk ) 21:12, 12 June 2011 (UTC) DGG ( talk ) 02:22, 20 June 2011 (UTC)
17. Oppose per Andy Slrubenstein | Talk 01:12, 17 June 2011 (UTC)
18. Oppose. per Slimvirgin. There is no problem to be fixed. The text says what it means, it means what it says, and it says it well. BE——Critical_{__Talk} 01:44, 18 June 2011 (UTC)
19. Oppose It misses an essential point, that for our collaborative Wikipedia project to work, editors must put aside thoughts about the truth of a matter and concentrate on what is in reliable sources. LK (talk) 06:29, 27 June 2011 (UTC)
20. Oppose — V = IR (Talk • Contribs) 05:09, 28 June 2011 (UTC)

Additional comments and discussions

Slim, that straw poll was farther reaching. A "yes" meant agreeing with two declarations, one that the (overall) first sentence is problematic, and second that it needs to be rewritten. And even then half of everybody said yes. And, at the time, without any specific proposal; this is a very mild one. Sincerely, North8000 (talk) 16:16, 10 June 2011 (UTC) Italic text

The poll asked whether the first sentence ought to be changed. It gained no consensus, and that's only on this page. You would need a strong consensus (significantly above two thirds) to remove "verifiability, not truth," which is a central idea in WP's policies, and you would need consensus far beyond this page. Lots of people have making that point for a few months, North. :) SlimVirgin ^{TALK|CONTRIBS}

Surely if it really is a central Wikipedia idea, it should be drawing in more than one-third support? Or to put it another way, if it only has the grudging support of about a half (or fewer) of the people who comment, it ought to be relegated from its position at the top of what is advertised as a key policy page?--Kotniski (talk) 17:30, 10 June 2011 (UTC)

The problem, as you know, is that a tiny number of you have been going on about this for months, and it wears most people down, who either don't bother to comment or comment once then wander off. That's why wide input is needed for fundamental changes to core policy, as a safeguard against the kind of thing that's been happening here. SlimVirgin ^{TALK|CONTRIBS} 17:36, 10 June 2011 (UTC)

It's hardly a fundamental change to policy - we all know what it's supposed to say (more or less), it would just be preferable to say it more clearly and accurately. Anyway, you seem to be voting oppose without any reason except that you expect people not to support it or you don't expect wide input, which seems rather premature (these are points to be made when the discussion comes to be closed). Do you have any argument for preferring the present wording - which tells people that we fundamentally don't care about the truth of what we write?--Kotniski (talk) 17:45, 10 June 2011 (UTC)

Slim, that's not what the straw poll question said if you read it closely. Second, I agree with Kotniski that this is not a change in policy, it's just a change in wording. If every change in wp:ver counts as change in policy, then you have changed the policy 5 times in the last 10 days. Third, you are bringing up that double standard again, but this even goes along with that, saying to cast a wide net for feedback and give time for lots of feedback. Sincerely, North8000 (talk) 18:02, 10 June 2011 (UTC)

You don't have to read the poll question closely, North. It asked whether the first sentence needed to be rewritten. It isn't reasonable to keep on ignoring people when they say no. SlimVirgin ^{TALK|CONTRIBS} 18:08, 10 June 2011 (UTC)

Well, it's written for all to see, and we are describing the questions and result very differently. I say that a "yes" also required agreeing that the current wording is problematic, you say not. I guess we need to agree to disagree and let folks read it for themselves. Second, I resent your characterization of this as "ignoring people when they say no" and feel that there is no basis for such a strong and nasty statement, or even a milder version of it. Sincerely, North8000 (talk) 18:55, 10 June 2011 (UTC)

SV we have a prime example of this nonsense festering its way though the santorum pages. One side has a bevy of sources using the the word and describing it as a neologism, whilst the expert opinion on the subject Partridge says it is NOT a neologism as it has not gained widespread usage. Which prevails RS or truth? NOT-TRUTH is fine for including articles on the paranormal and rejecting hearsay, it is not so good when it is being used to push an agenda. John lilburne (talk) 19:37, 12 June 2011 (UTC)

• Doesn't go far enough for me. A corrected first sentence should replace the word "threshold".—S Marshall T/C 19:36, 10 June 2011 (UTC)

So this is a baby step in that direction.  :-) North8000 (talk) 20:02, 10 June 2011 (UTC)

I'm glad you've confirmed that you're trying to change things significantly bit by bit, so the lobster doesn't notice he's being boiled. SlimVirgin ^{TALK|CONTRIBS} 20:09, 10 June 2011 (UTC)

I meant that as a way for S Marshall to potentially view this with respect to their comment. Further changes on this part would not be on my radar screen. Sincerely, North8000 (talk) 20:38, 10 June 2011 (UTC)

It's the third time in the last few weeks I've seen you use the term "baby steps" to try to persuade people to support your changes to the policy. SlimVirgin ^{TALK|CONTRIBS} 20:42, 10 June 2011 (UTC)

I consider taking a small safe step to be a good concept. I had a proposal in the list but this one was somebody else's. Sincerely, North8000 (talk) 20:53, 10 June 2011 (UTC)

However, this is a huge, unsafe step. Jayjg ^(talk) 04:28, 12 June 2011 (UTC)

Where's Crum? Short Brigade Harvester Boris (talk) 04:34, 12 June 2011 (UTC)

Can someone point to evidence that the current wording is actually causing confusion? I'm not really convinced there's a problem that needs to be solved. Mlm42 (talk) 19:56, 11 June 2011 (UTC)

IMHO it's pervasive in WP that once a situation starts getting contentious and moves into wikilawyer warfare, the idea of striving for accuracy becomes totally rejectable. And people keep pushing and mistakenly getting the impression that the Wikipedia/Wikimedia mission rejects the idea of striving for accuracy. This is NOT true, that impression comes from the faulty concept of trying to reverse engineer a mission statement out of imperfectly worded policies, which is backwards, because policies are supposed to implement the mission on objectives, not define them. And this sentence which for some inexplicable reason feels the need to put a "not" statement into a statement of what IS required has contributed to that. And so IMHO every one of the zillions of statements when someone discounts the idea / goal of striving for accuracy is an example of a problem that this sentence contributes to. Sincerely, North8000 (talk) 20:11, 11 June 2011 (UTC)

Well. Ok, but in terms of the wording, many of us simply do not see the problem you see, and these discussions have been going on for quite a while now. There's nothing wrong with the position that if there's no problem to fix, leave it alone. I just don't see any consensus developing here. --Nuujinn (talk) 20:41, 11 June 2011 (UTC)

So... you don't have any evidence it's actually causing confusion? Jayjg ^(talk) 04:28, 12 June 2011 (UTC)

To me, the proposed rewording has an identical meaning to the current version. The main difference between the two is that the current version has more emphasis placed on the contrast between verifiability and truth.. and I think it's important to emphasize that "verifiability" and "truth" are distinct concepts. The current wording does that well, as soon as possible. Mlm42 (talk) 01:35, 12 June 2011 (UTC)

On another point, I'm a little concerned that the line "whether readers can check that material in Wikipedia has already been published" is going to be misinterpreted as requiring free, online sources, or as requiring inline citations for everything (you know, because none of our readers know how to ask Mr Google whether a source exists). I'm not actually convinced that the primary purpose of citing our sources is to let the readers check anything. I think it's primary purpose is to help editors check the material. WhatamIdoing (talk) 02:19, 12 June 2011 (UTC)

That's not a change; that exact wording is in the current first sentence of the policy. North8000 (talk) 11:05, 12 June 2011 (UTC)

True, but it still worries me. We've recently seen people fussing about non-English sources because they can't check the sources, and the demands for free, online sources have been so persistent that it's enshrined on the list. IMO "users" would be more appropriate than "readers", if we're going to have this. WhatamIdoing (talk) 21:54, 12 June 2011 (UTC)

Reading the comments of those who support the proposal, and trying to see where they are coming from, I think part of the disagreement comes from the very reasonable concern that Wikipedia should not be implying that "not truth" is actually something we want. We want the truth as reliable sources see it, but not as Wikipedia editors see it. Consequently, what Wikipedia editors are tasked with doing is to find "verifiability, not truth", but this is done in search for what will be true. As many of us who oppose the proposal have said, if we ditch the "not truth" wording, we will open a floodgate of editors who want to push their versions of The Truth. Instead, would some sort of clarifying sentence, added after, be the way to make clear that we do not mean that we want "not truth"? I haven't thought through how to word it, but I think clarification may be helpful here. --Tryptofish (talk) 17:44, 12 June 2011 (UTC)

Seems reasonable to me. --Nuujinn (talk) 19:40, 12 June 2011 (UTC)

Seems reasonable to me as well and a possible middle ground.....where does that leave us on this? Wish you were there when we were developing/vetting proposals. Sincerely, North8000 (talk) 11:09, 13 June 2011 (UTC)

The main issue with the current wording of the sentence and the better wording of this proposal, with the use of the word "assertions", is that the "iconic" statement seems to be saying that Wikipedia is fine with publishing lies, just so long as they are lies obtained from other people. I would hope that this is certainly not true. There is a reason why the term Wikiality was ever created in the first place, even if it was meant as a joke. We need to have an explanatory first sentence that explains what we mean by verifiability. We don't need a catchy little statement that is often abused and mocked for its ridiculousness. The question is, sure the statement Verifiability, not truth may be iconic, but do we really want to be iconic for a badly worded, mocked statement? Silverseren^C 21:26, 12 June 2011 (UTC)

Policy by mockery? We shouldn't make policy based on whether mockery exists, but we should consider whether the mockery is based on a valid criticism. --Tryptofish (talk) 21:54, 12 June 2011 (UTC)

FWIW, I love that catchy little statement, both in terms of the rhetoric and for the bulwark it provides against those who know the Truth. It's like WP's little black dress. I suppose I have read too much pragmatism. --Nuujinn (talk) 22:03, 12 June 2011 (UTC)

Obviously we shouldn't be making policy based on mockery, but it is based on a valid point. As has been pointed out before, just look at WP:OTTO. If we truly followed through with the Verifiability, not truth statement, then of course we should make that an article, since it's verified, even if it is an absolute and utter lie. Really, if we have to have some sort of statement, Truth through Verifiability would be much more conducive toward what this encyclopedia is trying to do. All in all though, I don't think we need catchy statements for everything, I think for this sort of policy, we need straightforward, explanatory sentences that fully explain how this policy should be used. Silverseren^C 23:14, 12 June 2011 (UTC)

Examples?

Unless I missed it, I haven't seen anyone provide a link to an argument that would have been avoided with the new wording? I think several editors (like me) haven't seen such an argument. The main case for the new wording appears to be that it will prevent some arguments from happening.. so I don't think it's too much to ask for links to a few good examples? Mlm42 (talk) 04:46, 13 June 2011 (UTC)

If one realizes that this one sentence is just contributory (not single-handedly holding the smoking gun) I could give you hundreds of examples, or start with a few. But it should be on the basis of the material hopefully being informative rather than for this User:North8000/Page2#Useful method to take a whack at any thought. Sincerely,North8000 (talk) 10:56, 13 June 2011 (UTC)

Just a few examples would be nice.. at the moment I'm taking it on faith that there exist any examples at all.. Mlm42 (talk) 15:33, 13 June 2011 (UTC)

" ... we do not especially care about truth, merely verifiability"

"Yes, the coverage may be wrong, but WP:V's instruction to aim for "verifiability, not truth" does not contain an exception for issues about which we assume to know the (sadly unverifiable) truth"

"It doesn't matter whether it's true or not. "Verifiability, not truth" is the policy."

" ... Wikipedia cares about verifiability not truth."

"I don't really care if any of this is true or not; Wikipedia is based on verifiability, not truth"

"Our policy has been explained to you: we want verifiability, not truth"

David Wilson (talk · cont) 15:12, 15 June 2011 (UTC)

Thanks; but to me, at first glance, all of these editors appear to be correctly applying the policy? If anything, these examples show how useful and effective the current wording is.. Mlm42 (talk) 15:35, 15 June 2011 (UTC)

I agree, and also note that verifiability policy is not the main problem in most of those cases--mostly the contentions is arising from poor sources, poor application of source, OR and bad behaviour in general. --Nuujinn (talk) 15:59, 15 June 2011 (UTC)

There still is a big problem here, let me give another example, you may need to browse around this diff to get the full picture here. Count Iblis (talk) 16:13, 15 June 2011 (UTC)

This example is as excellent as it is scary. --213.196.218.59 (talk) 16:29, 15 June 2011 (UTC)

I'm afraid I can't see how blatantly inconsistent misstatements of the policy can be an illustration of how useful and effective its wording is. All of the above statements appear to me to imply that the people making them think that it's permissible to have untrue statements in Wikipedia articles as long as they're "verifiable". But that's absurd—if something is truly verifiable it can't possibly be false.

In the last four examples, the editors making the statements are very commendably arguing for the exclusion of unsourced or poorly sourced material. However there's no need to resort to patently fallacious statements to do this. All that's necessary is to point out that policy requires the challenged material to be supported with a citation to a reliable source. There's nothing wrong with also pointing out that the truth of the disputed material is not sufficient for it to be included, but to say that it doesn't matter whether the material is true or not as long as it's "verifiable" is simply nonsense.

|n the first two examples, the editors making the statements are arguing for the inclusion of material on the grounds that it's supposedly reliably sourced. I didn't check what the arguments of those challenging the material were for excluding it, but for the purposes of what these examples were intended to illustrate it doesn't really matter. To make one's case for inclusion it's necessary to at least engage and refute the arguments of those who have challenged it. An argument that we're only aiming for "verifiabity" rather than truth can't do that since it's patently fallacious.

David Wilson (talk · cont) 17:39, 15 June 2011 (UTC)

If something is truly verifiable it can't possibly be false. Cue the philosophical discussion! :-) It's worth pointing out that the term "verifiable" in Wikipedia jargon may differ from other definitions. If something is "verifiable" (in Wikipedia jargon), then it's conceivable that it's still false.. after all, reliable sources aren't always right. Anyway, you appear to be disagreeing with the meaning behind the WP:V policy, not merely the choice of wording. Mlm42 (talk) 18:19, 15 June 2011 (UTC)

You are basically agreeing that the examples are valid......that the wording makes the policy get misapplied/misquoted. North8000 (talk) 18:22, 15 June 2011 (UTC)

Am I missing something? I think the policy is being applied correctly in these examples.. Mlm42 (talk) 18:28, 15 June 2011 (UTC)

What I meant was: "All of the above statements appear to me to imply that the people making them think that it's permissible to have untrue statements in Wikipedia articles as long as they're "verifiable"." are examples of the effects of the problematic wording. Sincerely, North8000 (talk) 19:32, 15 June 2011 (UTC)

It depends on what you mean by "permissible" and by "untrue statements". For example, if there is reason to doubt a sourced statement, then one way to hedge is to say "Source X said that Y is true". Even if some editors "are 100% sure" that Y is false, that alone is not grounds for removal (maybe this is the point you are disagreeing with?). Specifying the source is one valid way to let an untrue statement into Wikipedia.. after all "Source X said Y is true" could be true without Y being true.

I'm still not convinced the wording is problematic.. the disagreements here appear to be with the meaning of the policy, rather than just the wording in the opening sentence. Mlm42 (talk) 20:25, 15 June 2011 (UTC)

I'm backing off from efforts to change the first sentence. I just wanted to not leave the examples section hanging.

On your last sentence, they always say "the test of how a law is written isn't what a good cop can do with it, it's what a bad cop can do with it. Since in Wikipedia, everybody is made a cop, so that test is even more appropriate here. North8000 (talk) 21:09, 15 June 2011 (UTC)

I have no idea where in this discussion / these discussions is the best place to jump in! So I'm just jumping in here ... When I came back to WP after a long break, I had real trouble with the idea of verifiability, not truth. Something had been written in a couple of articles which, although the information came from an obviously 'reliable researcher' (and I'm talking about world-renowned in their field), may not have been the whole truth. There were mutterings from other people who were concerned that the researcher may have wandered into a simple error, and (to my view) those 'mutterings' had a very great deal of validity. But - I couldn't include them in WP, because as at that time - and to the best of my knowledge, still - the places where the mutterings occured weren't as 'reliable' as the published scientific paper. Even though one could point out the obvious, from photographs, to do that fell foul of the WP:OR rule. It's likely that the 'mutterers' are speaking truth - very likely indeed - but as yet WP can publish nothing that challenges the original researcher's published paper, no matter how true the challenge may be. It was very hard for me to understand this;; so I'd be all for some kind of change in the wording. I simply have no idea what that change would be. Pesky (talk …stalk!) 04:23, 5 July 2011 (UTC)

A possible middle ground?

Well this has been out three days. As indicated, the "wide net" includes a substantial time frame for folks who are not the regulars here to have a chance to find out about this and weigh in; I imagine we get the regulars quicker and the non-regulars later. There is a common theme from each of the "sides", on the support side it's that we don't want policy to say we want "not truth", and the "oppose" folks say that phrase is an embedded and useful way to reinforce the verifiability requirement.

One of the "oppose" folks (Tryptofish, 17:44 12 June in discussion section), endorsed by another of the "oppose" folks (Nuujinn) put forth the idea of keeping that phrase, immediately followed by a statement (my interpretation) that such does not mean that Wikipedia does not seek accuracy. Should we consider this possibility as a possible "middle ground" which could also provide a long term resolution of this? Sincerely, North8000 (talk) 15:48, 13 June 2011 (UTC)

Thanks. I note that, below, there is opposition to new proposals, so I certainly don't want to tread on that. But, whenever editors here feel ready, it might be useful to discuss just one sentence to add, without taking anything existing away. --Tryptofish (talk) 21:29, 13 June 2011 (UTC)

Well we went through a full process to get proposals down to one before going out for this, just to avoid such complexities. I was thinking of it as more of a re-direction of the above process. I guess the proposal could be to leave the sentence as is and immediately after it something like "However, accuracy IS a goal of Wikipedia". I was thinking that if a few of the ardent folks from both sides say they like this, we'd float a proposal to both put this in and abort the main RFC. Sincerely, North8000 (talk) 21:48, 13 June 2011 (UTC)

North, this has become disruptive. You can't keep filing RfCs, abort them when they don't go your way, file a new one, rinse and repeat. The community supports "verifiability, not truth." Even if everyone on this page were to vote to remove it, it would be meaningless, just as a small group couldn't vote on the NPOV page to remove that WP's articles must be neutral.

Please do as people are asking on this page, and either let the issue drop or open a subpage for further discussion. But we need several months of no more polls or proposals about the first sentence (or second sentence or footnotes that might dilute it, etc etc). This talk page is for discussing the meaning of the policy, usually for the benefit of editors who arrive here with questions. That has barely been possible recently because the page has essentially been hijacked. SlimVirgin ^{TALK|CONTRIBS} 21:57, 13 June 2011 (UTC)

What I was talking about was based on keeping "verifiability, not truth". --Tryptofish (talk) 22:10, 13 June 2011 (UTC)

Slim, I consider your characterization of the situation and my actions to be nasty and inaccurate. We're had a lot of chaos on this topic over the last several months. All that I did was propose and assist a process to methodically and calmly bring it to a real conclusion, one way or another. That was for people to brainstorm and list proposals, then pick ONE which would go to this stage, and then cast wide enough of a net to get it settled either way. I had a proposal in which did not make the final cut; the main RFC proposal IS NOT MINE, it was somebody else's that emerged from that process. Now I was raising the idea of using Tryptofish's proposal as a compromise. North8000 (talk) 22:22, 13 June 2011 (UTC)

It has been brought to a conclusion: there is no consensus for these changes. Maybe revisit the issue in another 6 months? A Quest For Knowledge (talk) 22:28, 13 June 2011 (UTC)

Folks were saying it needs a wide net, which is both time and places. It's only been three days. I think that we should make sure we do that. I just hope that it either clearly passes or clearly fails, I personally don't want to revisit it even in 6 months. Since the process was organized & methodical (unlike the previous random stuff) if we let it play out, I think that most folks on both sides of the issue would be more likely consider it settled under a good process. Sincerely, North8000 (talk) 22:39, 13 June 2011 (UTC)

Fine, you have started an RfC, so please let it complete itself without trying to ask yet another question. SlimVirgin ^{TALK|CONTRIBS} 22:55, 13 June 2011 (UTC)

Conclusion

It has been about 3 days since the last vote. I think that the process was methodical and proper-looking enough so that it addresses both the specific proposal and the question of changing the core wording of the first sentence. With a near-dead heat by count (17 support, 18 oppose) this means that there is no consensus to change it, and no consensus to keep as-is. But I think that this means that the status quo prevails. As the one who sort of herded the process leading to this, this was disappointing to me, but also reassuring that it was arrived at via a process that was sufficiently broad and methodical, and I think without significant flaws or bias. Sincerely, North8000 (talk) 12:12, 15 June 2011 (UTC)

Proposal for a change in the first sentence #2

Failed proposal. causa sui (talk) 20:01, 13 July 2011 (UTC)
The following discussion has been closed. Please do not modify it.
{{rfctag|policy}} Another proposal for re-writing the beginning of this policy. 22:50, 12 June 2011 (UTC) Proposal Replace the first paragraph with: All material included in Wikipedia must be able to be verified in a published, reliable source. It is not sufficient for information to be true; editors must be able to check that a reputable source says it is true. Verifiability is a threshold for inclusion--a necessary condition which must be met before other considerations come into play. Not every edit which is verifiable must be included; other considerations regarding style and neutrality—length, relevance, weight, point of view, availability of better sources, notability of the subject, and editorial discretion—are important. Editors should consider all aspects of a source and its context before using it, so long as they do not engage in original research when doing so. Support Support as nom. I prefer it since I think it says what we mean, reflects actual practice, acknowledges the importance of context and editorial discretion, and is explanatory without being ambiguous. Is there support for such an approach? What are the drawbacks to going this way? Is it an improvement over the current version? I think there is interest in this approach. The drawback is slightly more length (3 sentences). For some, explicitly acknowledging editorial discretion invoke's chaos, but it means considering specifics, style guidelines, and neutrality issues within the bounds of original research. Which is what we regularly already do. I like the link to both NPOV and OR enhances the 'harmony' of the policies. Ocaasi t | c 22:50, 12 June 2011 (UTC) Support I liked this at the time it was proposed. It spells out what we mean. --JN466 21:07, 13 June 2011 (UTC) Support But I think that the discussion has sort of temporarily run out of gas. North8000 (talk) 18:01, 19 June 2011 (UTC) Support : Now it says that truth is a prerequisite, that style matters, that a source may have different interpretations if not exactly copied and i agree. Alkis0 (talk • contribs) 18:30, 30 June 2011 (UTC) Oppose Ocaasi, I accept your intentions are good, but the backdrop to this has been wall-to-wall polls and proposals since April—three already on his page, one just a few days old—several of them put forward by editors who do very little content editing. See here for a list. The result is that people are really fed up with it, and so just a tiny number of editors are left hammering away, which is the very definition of a lack of consensus. I think if people want to continue to discuss the first sentence, they should set up a subpage, so that this talk page isn't constantly being hijacked by these polls. SlimVirgin TALK|CONTRIBS 22:58, 12 June 2011 (UTC) SV, two considerations. There was a poll about changing the first sentence in general, then one of two ideas which received some support had a poll. This is the second one (proposal 43). I'm not sure a sub-page is really needed; if this page isn't for discussing changes to the policy, I don't know what it's here for. We keep getting in these disputes, not because people are misguided, but because the policy includes a mix of pith and ambiguity. Why not spell it out, say what we mean? That said, I understand these polls are getting old. I don't have more up my sleeve, but I do see the same issue with WP:V being a bit vague for new editors, and not well integrated with the other policies. So, I'd like to hear what editors think about this proposal. Ocaasi t | c 23:07, 12 June 2011 (UTC) But people said no to changing the first sentence, and some others ignored that and went ahead with other polls, which continue. The fact that we have proposal 43 tells you something. "Verifiability, not truth" is iconic. It can't be changed by a small group on this page. It has been extremely helpful for years in introducing new editors to the culture. They see that phrase and they suddenly get it. It prevents any group from taking over an article by claiming exclusive access to The Truth. And it reminds us that in many ways we're glorified (and perhaps not even glorified) stenographers. NPOV and "verifiability, not truth" are the backbones of Wikipedia; or as other editors have said, its secret sauce and little black dress. SlimVirgin TALK|CONTRIBS 23:15, 12 June 2011 (UTC) I absolutely agree that the counterintuitive "verifiability, not truth" statement is supremely able to make editors "get it". I "got it" by inwardly rebelling against that statement when I first started editing, then thinking about it, and then suddenly seeing what it meant. I am just concerned that editors these days sometimes use it as a justification to cut short discussion and include material, especially BLP material, that is not reliable, arguing that it does not matter if it is true, because it is verifiable (= someone else has printed it first). Perhaps what we really should be doing is to highlight that error in thinking, and add something to the policy that warns editors not to fall into that trap. "Verifiability, not truth", correctly understood, means "Your knowing it is true isn't enough to put it in Wikipedia." It should not be interpreted to mean, "We don't care if it is true or not, and you should not either. It does not have to be true for us to assert it as truth. It only needs to have been published." --JN466 21:25, 13 June 2011 (UTC) Oppose, SV has laid out the issue nicely. My suggestion would be for those who have issues with the current wording pick up the notion of working up an essay on the topic. --Nuujinn (talk) 10:22, 13 June 2011 (UTC) Oppose - This is beginning to become disruptive. How many polls do we have to conduct on this? The horse is dead... stop beating it! It should be obvious by now that any proposal that omits the "Verifiablility, not Truth" language will not gain a consensus. Blueboar (talk) 12:45, 13 June 2011 (UTC) If the first sentence of this proposal was "The threshold for inclusion in Wikipedia is Verifiability, not Truth" and then its meaning was actually explained as above, you would consider it? Ocaasi t | c 15:06, 13 June 2011 (UTC) Not really... As I have said in the previous six or so polls... I prefer the current language. I do understand the concerns that underlie all this, but I don't see anything that has been proposed so far as being an improvement on what we already have. Blueboar (talk) 15:26, 13 June 2011 (UTC) Oppose. Dear god, not again. I'm against this for all the reasons already stated in all the other previous discussions. Concur with Blueboar that these repeated polls are starting to become disruptive. A Quest For Knowledge (talk) 12:56, 13 June 2011 (UTC) Oppose per User:SlimVirgin. It is an absolutely fundamental mantra of Wikipedia, Hell will freeze over before we abandon it. Roger (talk) 18:21, 13 June 2011 (UTC) Oppose. WP:LETITGO. —chaos5023 (talk) 21:39, 14 June 2011 (UTC) Polls are evil, but I agree with the previous two comments. --causa sui (talk) 19:40, 16 June 2011 (UTC) Oppose, I think. Mine eyes are getting bleary and these are looking all alike; over and over again. Unlike the powerful and clear statement "Verifiability, not truth!". Dreadstar ☥ 03:21, 17 June 2011 (UTC) Oppose per what Slimvirgin said. LK (talk) 06:31, 27 June 2011 (UTC) Strongly oppose. per Roger above. Wtmitchell (talk) (earlier Boracay Bill) 01:11, 28 June 2011 (UTC)

Threaded discussion

• It doesn't seem like much of an improvement to me. It perpetuates many of the flaws in the current version—the poor writing, the unnecessary circumlocution, the wordy passive hortatives in place of simple active imperatives, the word "threshold" in legalese (you might as well say "whereas" and "hereinbefore"), and in fact all the problems associated with the way this policy has grown: written by a committee with widely differing agendas about its function and purpose, developed by stealth-edits that are subsequently declared unchallengeable, revised in the light of hard cases, and now followed unthinkingly by the unthinking. You could fix it to an extent just by simplifying the circumlocutions ("be able to be verified"→"be verifiable"; "not every edit which is verifiable"→"not every verifiable edit"; "editors should consider"→"please consider", and so on). But unfortunately, sheer RFC-fatigue is going to mean the Anti-Change Party wins this one. I think that realistically, there's no way to remove the worst phrases from this policy for another few months, and any change will have to be accomplished by stealth-edits, the way so much of this policy was written in the first place.—S Marshall T/C 23:18, 12 June 2011 (UTC)

It's an unfortunate result of the double standard.North8000 (talk) 11:13, 13 June 2011 (UTC)

Are these expressions of Good Faith? --Nuujinn (talk) 14:44, 13 June 2011 (UTC)

All material included in Wikipedia must be able to be verified in a published, reliable source.

The problem here is that you often can't verify statments from single sources. So, "a published, reliable source", should be replaced by "published, reliable sources"

It is not sufficient for information to be true; editors must be able to check that a reputable source says it is true.

Here the problem is that verifiability does not mean that you can find a reputable source that explicitely says that a statement is true. What often happens is that the statement can be verfied from an entire body of literature. This is is often the case for scientific topics. It's not for nothing that it takes some years of study at university to become a scientist. Uncontroversial accepted facts are often not verifiable by direct citations to a source where the fact is presented verbatim.

Verifiability is a threshold for inclusion--a necessary condition which must be met before other considerations come into play. Not every edit which is verifiable must be included; other considerations regarding style and neutrality—length, relevance, weight, point of view, availability of better sources, notability of the subject, and editorial discretion—are important. Editors should consider all aspects of a source and its context before using it, so long as they do not engage in original research when doing so.

This looks ok. but only with the more liberal definition of verifiability. Count Iblis (talk) 22:14, 13 June 2011 (UTC)

The RFC bot is choking because there's more than one open RFC on this page. This one isn't being listed (it's listing the other twice instead), so I'm disabling the RFC tag in the hope that the other one will be correctly listed (once, and with the text rather than just the link). Keeping this one "open" isn't having the desired effect of getting it listed, so turning it off isn't going to make this any less invisible than it already is.

Harej would be happy to hear from anyone who wants to help identify and fix the coding problem; see WT:RFC. WhatamIdoing (talk) 21:31, 8 July 2011 (UTC)

Let's verify that "verifiable" is intimately related to "truth"

See here:

verify early 14c., from O.Fr. verifier, from M.L. verificare "make true," from L. verus "true" (see very) + root of facere "to make" (see factitious).

And here we see that:

–verb (used with object), -fied, -fy·ing.

1. to prove the truth of, as by evidence or testimony; confirm; substantiate: Events verified his prediction. 2. to ascertain the truth or correctness of, as by examination, research, or comparison: to verify a spelling. 3.

to act as ultimate proof or evidence of; serve to confirm.

Count Iblis (talk) 15:56, 13 July 2011 (UTC)

We use the word in the more legalistic context of "confirmation" ... closer to the third definition you list... "serving to confirm".

Verifiability (on Wikipedia) means that we are able to confirm that something is stated by a reliable source... however, it does not necessarily mean that what is stated by that source is "true". And we definitely don't accept an assertion of "truth" as verifiability. Blueboar (talk) 16:27, 13 July 2011 (UTC)

And we accomplish 2 by verifying assertions with reliable sources, hence verifiability trumps truth in practice. --Nuujinn (talk) 19:28, 13 July 2011 (UTC)

• Nuujinn and Blueboar, it really does seem to me that you're getting around the word "truth" by using synonyms for "truth". I think you're inserting a semantic layer between "reliably-sourced" and "true", and then mistaking the semantic layer for something that's actually meaningful. But I think these verbal strategies that circumnavigate the word "truth" are unsuccessful, because a source is only "reliable" because it's "accurate", or "widely-accepted" (i.e. "probably true".) Reliability comes from "editorial supervision" and "a reputation for fact checking" (i.e. "more likely to be true.") Everything that makes a source more acceptable on Wikipedia is also, inevitably, something that makes it more likely to be true. Do you see?—S Marshall T/C 20:30, 13 July 2011 (UTC)

I don't think that we're dodging the word "truth". I think that we're acknowledging that what we write might be wrong, and that even if we know and can personally prove, beyond any shadow of doubt, that what we believe is 100% true, that material may not be appropriate for Wikipedia. Two examples might help:

• I'm wearing a red shirt today. This is 100% true. There is no question of the accuracy of this statement. I could prove it to anyone's satisfaction via photographs and sworn witness statements. But it's not verifiable, so we aren't going to include this. "Being true" is not sufficient. We do not include material because it is true.
• If Wikipedia were written a thousand years ago, we'd have written that the Sun revolves around the Earth. This is false—but we would have said that. "Being true" is not required. "Being published by reliable sources" is. Given what we know about the accuracy of our sources (you know what they say about newspapers? The 90% you don't know anything about is absolute truth, but the 10% you have personal experience with is completely wrong), I am convinced that Wikipedia contains many outright falsehoods. We're okay with this: we'd rather have a verifiable falsehood than an unverifiable truth. Ideally—and usually—what is verifiable will also be true. But when that ideal state is not available, we choose verifiable lies over unverifiable (alleged) truth. WhatamIdoing (talk) 22:16, 13 July 2011 (UTC)

• I'm perplexed by this angle, WhatamIdoing. For the avoidance of doubt, my position is that verifiability is a criterion for inclusion on Wikipedia. It's not the only criterion, which is why your red shirt article wouldn't make it in. As you say, being true is not sufficient and we do not include material merely because it's true. But that's entirely tangential to the point I was making.

  Equally, your remarks about our current state of knowledge are also totally accurate—and indeed, you're echoing a remark of mine which User:Peter jackson chose to enshrine as Raul's 301st law. But it's also entirely tangential to the point I was making.

  Will it be helpful if I explain my position again in different words?—S Marshall T/C 22:39, 13 July 2011 (UTC)

Probably not, but go ahead anyway. Blueboar (talk) 23:23, 13 July 2011 (UTC)

I think that the "we don't care if it's false" type of statements are mistakenly created by trying to reverse engineer mission statements from policies and guidelines instead of the reverse which is how it should be done. Sincerely, North8000 (talk) 00:04, 14 July 2011 (UTC)

@S Marshall, just to be clear, I'm arguing about what truth is from a philosophical position learned during graduate school where I concentrated on literary theory and philosophy of language. My personal takes on this are mostly informed by Pierce and Rorty, with some Wittgenstein and european phenomenology thrown in. Many of the arguments above seems largely informed by common sense, logic, science and math, which are equally valid approaches. My experience is that defining truth is tricky, and how one makes a determination of what is true and what is not is generally determined by context. Here, we have an operational definition, where we say that we determine what is accurate and true via verifiability, which is a good way to do it because we can come to consensus about what are good sources and what is verifiable. As you can see from the discussion above, we have more trouble defining truth. And if you are familiar with Kuhn, you know that science itself shifts widely from time to time regarding what is true, also based on what can be verified via experimentation. Religion generally defines truth via dogma, but that varies widely. So basically, my take is that in any context, what is true and what is not is determined by an accepted procedure. Where people disagree about the procedure, agreement about truth is not possible, for example, evolutionist versus creationists. So I see a very significant advantage in focussing on the procedural definition, verifiability, and not getting bogged down in discussion about what it true and what is not true. I hope that's clear. --Nuujinn (talk) 00:44, 14 July 2011 (UTC)

Nuujinn, is that the big departure that it appears? .. you appear to be saying that meeting wp:verifiability defines it as true? North8000 (talk) 01:07, 14 July 2011 (UTC)

Time for a little forward progress I don't think that anybody is challenging the following statement: verifiability is a requirement for inclusion. Nothing else, such as truth, is a substitute. So, should we stop "pretending" that that is being debated? North8000 (talk) 01:11, 14 July 2011 (UTC)

The real debate is that should we have wording (or lack of wording) that gets construed as saying that a quest for accuracy is not another legitimate editorial objective?...given that it does NOT override the verifiability requirement? North8000 (talk) 01:16, 14 July 2011 (UTC)

I would hedge that... what we strive for is an accurate presentation of what is said by the various reliable sources that discuss a given topic. This means that when sources disagree (and they often do), we note the disagreement and include discussion of minority views (in accordance with WP:DUE, of course). We don't omit significant minority views, even when we personally believe that the view expressed by the source is wrong. to do this would not accurately present what the sources say. Blueboar (talk) 02:20, 14 July 2011 (UTC)

Exactly. North8000, I am saying that the only way we could possible determine what is true in discussions here, as a practical matter, is by an appeal to reliable sources, and we have a procedure for that. We do not agree about the nature of truth, and historically, truth doesn't cross disciplines well. And I feel strongly enough about that that I do not support removing "verifiability, not truth", period. I am happy to discussion improving the explanation of that phrase, but not changing or removing it. --Nuujinn (talk) 02:39, 14 July 2011 (UTC)

Agreed.
— V = IR (Talk • Contribs) 03:12, 14 July 2011 (UTC)

Yes. I also feel strongly that the specific language "verifiability, not truth" is a crucial cognitive shock serving the indispensable function of signaling to new Wikipedia editors that the assumptions they came in with are all going to need to be re-examined. I oppose removing it and anticipate continuing to do so. In closing, dude, WP:LETITGO. —chaos5023 (talk) 03:21, 14 July 2011 (UTC)

I still like "Not only true, but verifiably true." We're never going to get 100% agreement on any of this stuff, simple because of the various reasons that have been outlined above; common misconceptions, 'grey area' truths, typos in combination with a paucity of reliable sources, and all that jazz. And we're always going to get the POV warriors, the WikiLawyers, and all the rest that muddies the field. Yes, we can invoke IAR on the odd occasion where something has obviously - or even almost certainly - sprung from a typo in an otherwise reliable source. But - talking of POV pushers and WikiLawyering ..... how about we don't join the ranks here in this discussion :o)? Any small step towards improvement in the ways policies are worded and applied is a small step in the right direction. Pesky (talk …stalk!) 08:04, 14 July 2011 (UTC)

Pesky, if you require the material to be "not only true", then you require an editor who believes that homeopathy works to remove all information that says homeopathy is pseudoscientific garbage.

There's no two ways about that: either you are required to include only information that you believe is true, or you are permitted to include information that you do not believe is true. This is an "A/not-A" choice. There is no middle ground.

Wikipedia actually wants people to include (and accept the inclusion of) information that they personally do not believe is true, merely because it is verifiably the position of a sizable body of sources (even if, in the opinion of editors, those sources are absolutely, provably, 100% wrong). We must include information that is verifiable-but-IMO-false. WhatamIdoing (talk) 17:30, 14 July 2011 (UTC)

• Nuujinn, it's not necessary to discuss the philosophy or linguistics of "truth" if you simply remove the word "truth" from the policy entirely. To be clear, that's what North and I have been banging on about all these months: to remove the words "not truth". There's no need to open that can of worms at all.

  When editors say "verifiability, not truth", what they mean is "verifiability, not truthiness". They don't literally mean that it's okay to add falsehoods to the encyclopaedia. That's what ought to be clear.

  "Verifiability, not truth" is a soundbite. It's popular with contributors to this page because it's helpful in dealing with POV warriors. The only problem with it is the one Hans Adler points out so clearly earlier on this very page, before we got another huge ForestFire of discussion spreading out in all directions (it would really help if one day we could have a discussion that doesn't lose focus, by the way).—S Marshall T/C 11:06, 14 July 2011 (UTC)

  It seems to me that what you're describing is perfectly clear to those who are willing to understand it (the vast majority). There doesn't appear to be a significant issue with this, and it's a pithy statement that is easy to remember and conveys an easily understood concept, so I don't understand what the problem is. "If it ain't broke, don't fix it."
  — V = IR (Talk • Contribs) 11:56, 14 July 2011 (UTC)

  The problem is that policy should document practice, and in practice, we don't follow "verifiability, not truth". This is what I was talking about further up the page, if you wouldn't mind reading back?—S Marshall T/C 16:10, 14 July 2011 (UTC)

  I didn't really find that argument compelling. While of course people are going to go by truth, there's a workable core policy for contentious cases (those being the only cases policy ever really needs to address) in 1) going by verifiability using reliable sources 2) determining what reliable sources are by consensus. Where your reasoning falls down is when you assert that we're deciding that sources are reliable because they tell the truth; while that may be the motivation for many or most of us, it's not the mechanism by which we make that determination; the mechanism is consensus. —chaos5023 (talk) 17:37, 14 July 2011 (UTC)

  Right, exactly. I'd be interested in your reply to this though, Marshall. I'm not quite sure where it is that we're not seeing eye-to-eye (and I've at least skimmed though, if not actually read, all the discussion above). In my experience we do follow "verifiability, not truth" in practice, so... uh, I don't accept that we're not already documenting practice.
  — V = IR (Talk • Contribs) 18:06, 14 July 2011 (UTC)

  Though this kind of gets to something that's been kicking around my head for a while, which is that there's an enormous disconnect between how Wikipedia operates (and should and must operate) in uncontentious areas and how Wikipedia operates (and should and must operate) in highly contentious areas -- but we act almost ashamed of this, like if we say it out loud then we're articulating a double standard, which I guess is inherently bad, undemocratic and anti-wiki-way. So we get policy that's mostly dedicated to nailing down what to do in the worst of cases -- which people then read, say "oh, I understand now", and proceed to devastate the countryside by applying it consistently. I'm kinda just thinking out loud here, but I think this is a problem. —chaos5023 (talk) 18:17, 14 July 2011 (UTC)

  Yea, I agree, that's something that I've come across as well. Then we end up dealing with what are essentially "policy warriors" running around for a while. I'm not sure that there's a way to resolve that issue, though.
  — V = IR (Talk • Contribs) 18:32, 14 July 2011 (UTC)

From my experience, "policy warriors" inevitably take bits of policy out of context, and (perhaps more importantly) focus on one policy to the exclusion of all the others (thus not seeing how they impact and influence each other). This is a flaw with the editors, not with the policy. The fact is, you can not fully understand what we mean by "The threshold for inclusion in Wikipedia is Verifiability, not truth" without also understanding what we mean by "Editing from a neutral point of view" and "Wikipedia articles must not contain original research"... and vise-versa. All three of our core policies are expressing the same basic concept... just from different angles. Blueboar (talk) 18:47, 14 July 2011 (UTC)

We can take (should take?) the line that (for example) "Proponents of homeopathy say that [blah blah blah]", and "Opponents of it say that [blah blah blah]" - both of which statements are verifiably true. We can also say stuff like "The theory of homeopathy is [blah blah blah]" and then say something like "Scientific experiments have shown that [blah blah blah]" - again, both statements verifiably true. This kind of approach can apply perfectly well to all such stuff, and pretty much any other area where POV-warriors start baring their teeth and flexing their claws. I actually think that this kind of approach is exactly what's meant by NPOV - we state the position of both sides but anyone reading it wouldn't be able to tell which side the writer was on (if they were on either). Pesky (talk …stalk!) 09:35, 15 July 2011 (UTC)

Not exactly; you're making a WP:GEVAL error. DUE requires us to present the dominant view as being the dominant view. So we do say "Proponents say that it works" but not so much "Opponents say that it doesn't". Instead, we say simply "It doesn't".

If I were a true believer in homeopathy, and Wikipedia's policies said that I must not leave false information in an article, I would feel compelled by the policy to delete the sentence "Homeopathy's efficacy is unsupported by the collective weight of modern scientific research" as IMO false and inaccurate. I would be convinced that you meant me to delete "The modern mechanism proposed by homeopaths, water memory, is considered implausible in that short-range order in water only persists for about 1 picosecond", as I would consider that an extremely plausible mechanism. I would definitely delete "The proposed rationale for these extreme dilutions – that the water contains the "memory" or "vibration" from the diluted ingredient – is counter to the laws of chemistry and physics, such as the law of mass action", since I would consider it a normal working of chemistry and physics.

Under the current system, I would leave these (IMO) "false" statements in the article, because Wikipedia says from the start that it doesn't care whether the material is truly accurate, so long as it accurately reproduces the (IMO) "errors" made by mainstream sources. I would agree that these errors are verifiable, and therefore eligible for inclusion under the current system. WhatamIdoing (talk) 16:40, 15 July 2011 (UTC)

Suggestion for any other changes

Per ThatPeskyCommoner, I think we've been at loggerheads dealing with "The threshold for inclusion in Wikipedia is verifiability, not truth" for months now, so let's try a different tack and leave that aside for a reasonable period (I'd say 90 days, but that's just my preference) and focus on the rest of the text. The second half part of the first sentence is "--whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true." What can we do to improve that clause? Do we change that to explain that we use the first clause as a kind of touchstone, or expand it to capture the relationship? Or something else? 10:39, 14 July 2011 (UTC)

I think that that second half of the sentence is fine. I'd suggest a new sentence after that to mitigate the unintended effects of "not truth". Maybe my awkwardly worded "This does not intend the discourage seeking accuracy; verifiability helps achieve accuracy" could be a starting point. North8000 (talk) 11:06, 14 July 2011 (UTC)

That sounds fine to me, fwiw. --Nuujinn (talk) 12:56, 14 July 2011 (UTC)

I would want to tweak the wording... but the concept is good. I could live with something like that. 13:13, 14 July 2011 (UTC)

Maybe just "Verifiability helps achieve accuracy". ? North8000 (talk) 13:28, 14 July 2011 (UTC)

I see where you're going with this, but most of the purpose of "verifiability, not truth" has to do with... avoiding "value judgements". I mean... as editors, we're supposed to be neutral. We're not supposed to be saying "this is what I learned, I know that it's true. This is what should be in the article because it's the truth." That sort of thing should be avoided in all cases (although we'll get there 99.999% of the time anyway just because that's what all of the sources are saying, it's then coming from the sources, not from any particular editors). So... I'm not sure why "a new sentence after that to mitigate the unintended effects of "not truth"." is something that is actually desirable. Actually, I'm somewhat suspicious that adding something there will only create confusion.
— V = IR (Talk • Contribs) 18:14, 14 July 2011 (UTC)

By neutral, do you mean no value judgemet as to whether to put in knowingly wrong or irrelevant sourced material vs. knowingly right or relevant sourced material? And, if so, respectfully, where did you get that from? North8000 (talk) 18:44, 14 July 2011 (UTC)

A comment which may or may not be useful:

Content must be verified per RS, and content must be accurate. For example, that someone is wearing a red short and that is verified creates accurate content. Truth is another issue altogether. Truth is an ultimate, and is also subjective. We shouldn't confuse accuracy with truth. Our article must contain it seems to me content that is both verifiable with the best sources and best editor oversight available to provide the most accurate article possible. Truthfulness doesn't really enter the picture.(olive (talk) 19:23, 14 July 2011 (UTC))

Yes! exactly. Thank you.
— V = IR (Talk • Contribs) 20:41, 14 July 2011 (UTC)

Object lesson

Here, maybe a firm example will help everyone understand each other better.

NASA, through the Apollo program, landed men on the Moon through the late 60's and early 70's. There's documented proof of this, and it's well accepted that it happened. However, there is a significant minority, world wide, who hold what is a fringe belief that the moon landings never happened. This is a high enough profile fringe theory that not only is it mentioned in the Apollo project article, but it has it's own complete article at Moon landing conspiracy theories. This is the type of article that not only benefits from the "verifiability, not truth" phrase, but it's almost dependent on it.

Wikipedia should not be advocating either for or against those conspiracy theories. That we cover these fringe theories at all can only happen because their existence and the details of their views is verifiable. We're not ignoring the truth, but... Climate change articles are another area that is basically dependent on this, as well. Some of the worst behavior issues that this policy addresses are those of well established editors rather than the fringe theorists or the POV pushers (who have to deal with "no original research" much more than verifiability. There are plenty of well-healed cranks who will publish fringe theory material, after all). Just because these ideas (such as "we didn't land on the Moon" or "CO2 doesn't cause global warming") are wrong doesn't mean that we shouldn't cover them. "Verifiability, not truth" allows for that, and allows the encyclopedia to (for the most part) maintain it's neutrality.
— V = IR (Talk • Contribs) 18:28, 14 July 2011 (UTC)

Both excellent examples. Blueboar (talk) 19:17, 14 July 2011 (UTC)

• Well, let's run with those examples and explore them in detail, because I think they support my case. There are sources for two views: 1) that the moon landings happened, and 2) that they're a hoax. There are more, and better, sources for view #1.

  Sure enough, Apollo program says that the moon landings happened. It doesn't say that they "allegedly happened". It compares the sources and then, in a series of simple, declarative sentences, takes one side over the other—and contrary to what Ohms Law said above, I was unable to find any mention of view #2 in that article whatsoever.

  Meanwhile, Moon landing conspiracy theories acknowledges that sources for view #2 exist, but in language that clearly distances itself from that view. Again, though there are two kinds of source, Wikipedia clearly decides between them and selects one.

  For the avoidance of doubt my position is that our practice in this is correct, in other words, that the articles we're considering are correctly written. I'm simply pointing out that we do not do what this policy says we're going to do. We compare the sources and decide which are more "accurate", more "reliable", more "trustworthy", or other synonyms for "true".—S Marshall T/C 20:05, 14 July 2011 (UTC)

Edit conflict, wrote this without seeing S Marshall's. ::I think that the Apollo example shows the reverse of what you intended. The editors said that we landed on the moon, they didn't just tally up sources and say that we did and didn't. Being a non-controversial article (for 99% of people) they were able to ignore the "not truth" mantra.

The global warming one is immensely complicated, (starting with "what is the question?") and there is no view which is such a tiny minority in sources that it could be called fringe. Sincerely, North8000 (talk) 20:09, 14 July 2011 (UTC)

It looks like the Moon landings articles have slid over the last year or so, unfortunately (I haven't actually worked on them in a while). The Climate Change area is... slightly better (if only because arbcom got involved there, and slapped a couple people upside the head over verifiability). But the reason that I bring up both of these cases is because it illustrates the problem with the approach that "truth is better" that both of you seem to be taking. I don't hold them up as good examples, but as problem areas where "verifiability, not truth" needs to be taken into account (Keep in mind that there's a constant push - pull with what is "undue weight" to consider, as well).
— V = IR (Talk • Contribs) 20:39, 14 July 2011 (UTC)

• Ohms Law, the climate change case supports my view. In fact, I specifically cited global warming as supporting my case in this diff, on this page, yesterday. Also, I suspect you may have misunderstood where North8000 and I are coming from when you describe our position as "truth is better". That isn't our position at all. Our position is that "truth" is a problematic concept, meaning as it does different things to different people, which is why we should remove all mention of truth from this policy. And the phrase "not truth" is specifically problematic because it's against the basic purpose of writing an encyclopaedia to publish something false.—S Marshall T/C 21:11, 14 July 2011 (UTC)

  Alright, but the end result is the same. I disagree with your position that all mention of truth should be removed, primarily because "seeking the truth" is the primary problem that we deal with. It's always been a core tenant of Wikipedia that we're about "verifiability, not truth", to the point that it's a mantra now, so... I guess that I don't think that it matters how much it's applied where it needs to be applied most, saying it is still important. The problem that the idea isn't applied as it should be in some cases doesn't mean that we should ditch it, it means that we should discuss the idea further. You guys seem to misunderstand a fundamental pillar of Wikipedia's editing policy, so a bit of education seems to be in order here, to me. I'd be just as much against ditching the idea of "no original research" (by redefining what original research means) as I am to ditching verifiability, which yourself and others are advocating for by seeking to redefine what verifiability means.
  — V = IR (Talk • Contribs) 00:39, 15 July 2011 (UTC)

• The idea that North8000 and I wish to "ditch verifiability" is a serious mischaracterisation of our good faith intentions. We want verifiability to be a policy, and we want it to say that everything challenged or likely to be challenged must be verifiable by means of an inline citation to a reliable source. Please desist from setting up straw men. The fact that we want to change two particular words that SlimVirgin added, apparently without discussion, in 2005, does not make us weird extremists who want to wreck the encyclopaedia's fundamental structures. Others, as well as you, have accused us of this and I'm growing a little tired of it now. Please stop.

  I also see an attempt to claim that North8000 and I don't understand the policy, which is a claim I've also seen before. That's rather insidious, because it sets you, Ohms law, up as some kind of authority figure or teacher who can "correct" our "misunderstanding". In fact, I think our position shows that North8000 and I have read the policy and found it ambiguous, and we wish, in entirely good faith, to correct it.—S Marshall T/C 07:37, 15 July 2011 (UTC)

Yes, both S Marshall and I have have made it abundantly clear many times over that we support verifiability. How could you possibly get that backwards and say the opposite? Please stop. North8000 (talk) 09:30, 15 July 2011 (UTC)

It's still not clear to me how changing the "Verifiability, not truth" wording would solve any problems. The moon landings example, like other conspiracy theories, is an interesting one. There are people who are convinced that the moon landings didn't happen. Our Wikipedia articles will naturally attract those people, who will come and potentially argue ad nauseam about what the "truth" is. "Verifiability, not truth", is a quick and easy way of dismissing these arguments, instead of wasting time engaging in the arguments. Editor time is valuable, and policies should make it easy for good editors to continue doing good work.

Otherwise, a good editor might be tempted to waste countless hours engaging the conspiracy theorists, and attempting to "prove" that they actually did land on the moon (personally, I'm with Buzz Aldrin, and would want to punch the conspiracy theorists in the face). The point is that the burden should be on the conspiracy theorist. Of course we strive for accuracy - that's obvious. Consider the statements: (1) "Astronauts landed on the moon. Some people think it was a hoax.", and (2) "NASA claims that astronauts landed on the moon, but others disagree." I believe both statements are true, but the first one better reflects the sources. Some people believe that only the second one is true. Fortunately, that doesn't matter, because Verifiability, not Truth, is what's important - so we go with the first statement, and avoid arguing with the conspiracy theorists. (This comment is made independent of what the actual articles currently look like.) Mlm42 (talk) 04:04, 15 July 2011 (UTC)

Errrrm ..... an editor who really understands NPOV writing would surely just act as a reporter of what the theories say, and what the opponents of the theories say? Our job, if we're basically quoting/ re-wording what other people say, is not to make value judgments on what they're saying - just to report it? Pesky (talk …stalk!) 09:40, 15 July 2011 (UTC)

Just adding - if I were to say (re the moon landing conspiracies) that "The moon landing conspiracy theories are X,Y and Z, and these are the grounds on which they base those theories [gives grounds, cited, sourced, etc.]". "Opponents of the conspiracy theories state a,b, and c, and these are the grounds on which they base those statements [gives grounds, cited, sourced etc.]" then I very much doubt whether anyone reading this would be able to hazard a guess as to "which side I'm on". I could do exactly the same with almost any other contentious area, and leave readers at a complete loss as to which version I personally 'prefer'. This is what we're supposed to do - present the evidence (being what each side says and why) in such a way that only the evidence itself can sway the reader one way or the other. This is not hard to do - though it can be hard to explain. Pesky (talk …stalk!) 09:54, 15 July 2011 (UTC)

So you would re-write the Apollo article to not simply say we landed on the moon? just say there are two points of view (we did and we didn't) and maybe that one is a majority view? North8000 (talk) 10:10, 15 July 2011 (UTC)

The Apollo article itself I would write as reported. I would include a section on the conspiracy theories, and give each point of view 'due weight' (trying to make sure that the coverage of each point of view reflected the percentage of sources which supported that view). Pesky (talk …stalk!) 10:40, 15 July 2011 (UTC)

I'm only persisting because we're getting to a core item. My question was about the core statements... would you remove statements which state as fact that we landed a human on the moon? North8000 (talk) 10:49, 15 July 2011 (UTC)

Personally, no I wouldn't, because of the weight of evidence (including 'scientific weight' and 'reliability' of evidence) points to it having happened. I would have a separate article on the conspiracy theories themselves, including the history of the theories, where they originated, by whom they were originated, and so on - and I'd personally have that as a 'see main' link from the moon landing article, so that people who really wanted the info on the conspiracies could then switch to that article to read up more about them. But that's just what I personally would do - if it's an 80/20 split on the moon landings, I'd weight the article to the 80, and include 20 on the conspiracy theories, with a link to the article on the conspiracy theories. Pesky (talk …stalk!) 11:03, 15 July 2011 (UTC)

I think this one is closer to 99/1. So the editors decided that the 99% viewpoint was accurate, and stated what was in those 99% sources as fact, and chose not to use what was in the 1% sources. So they followed both wp:ver, and also exercised judgement in a quest for accuracy. They followed the policies, although they did not follow the pervasive false mantras mis-derived from policies. North8000 (talk) 12:30, 15 July 2011 (UTC)

Pesky's point about Due Weight is the key here... Due weight is an important consideration, not only in how much space we give to various viewpoints, but also in how we present them. We usually present viewpoints that are accepted by significant majorities of reliable sources as being "fact"... we usually present viewpoints with noteworthy minority support as being "Opinion". If the balance of reliable sources expressing the various viewpoints is more evenly split, then we present all viewpoints as being opinion (even the one that has a majority).

Assessing Due Weight is a function of WP:NPOV, and not a function of WP:V... with one important exception: If a viewpoint is not verifiable then we give it no weight at all (i.e. we do not include it... no matter how true we think it may be). That's what we mean by: "The threshold for inclusion is Verifiability, not Truth". Blueboar (talk) 13:01, 15 July 2011 (UTC)

You said that much better than I did! Pesky (talk …stalk!) 13:12, 15 July 2011 (UTC)

I agree, well formulated indeed. Also with regard to the discussion concerning fringe and POV-pushers above it is important to note that those guys tend to attack the "not truth but verifibability"-line from both sides. If we put the focus on "truth" (verifiability as "best" approximation of it), they will argue their POV/fringe based on their (fringe) "truth". But if we focus on verifiábility only they will argue their POV/fringe based on (false) sources claiming their "truth" is verifiable. So there is no easy way out here. Though on the verifiability side we can restrain the abuse somewhat by requiring sources to have a high degree of reputability (and accuracy).--Kmhkmh (talk) 13:38, 15 July 2011 (UTC)

We're agreeing except that you are avoiding my point: Would you say that the objective (in this case, where the subject is clearly a matter of fact) of all of what you describe is to put accurate information in Wikipedia, or is that process itself the objective? North8000 (talk) 13:35, 15 July 2011 (UTC)

If it's 99/1, we weight the article 99/1. The article wouldn't be complete if it had no reference at all to the conspiracy theories. Pesky (talk …stalk!) 14:02, 15 July 2011 (UTC)

(ec) Yes and no... our objective is to accurately present what the sources say... but it is not our objective to determine whether what the sources say is accurate or not. Blueboar (talk) 14:09, 15 July 2011 (UTC)

But, as exhaustively explained with examples such as Apollo program above, we do make value judgments about sources.—S Marshall T/C 14:21, 15 July 2011 (UTC)

Blueboar, we're getting to the heart of it. Your statement "our objective is to accurately present what the sources say" if taken to mean everything that it could cover, goes far beyond wp:ver which just specifies a criteria for inclusion of material. Where did you get that from? North8000 (talk) 14:32, 15 July 2011 (UTC)

@blueboar: Actually to some degree with we do determine (or better maby care about) whether a source is accurate or not: We do that via source selection, i.e. directly through editorial decision not to use an inaccurate source and indirectly through reputability as a filer or proxy (unreputable sources get ignored). The determination of a source's accuracy is part of an editor's source slection process.--Kmhkmh (talk) 14:34, 15 July 2011 (UTC)

Perhaps... but this is a function of reliability and due weight not verifiability. Let me rephrase what I said above... our objective is to accurately present information related to our subjects... this can mean that we must present information we may personally think is inaccurate. Blueboar (talk) 15:12, 15 July 2011 (UTC)

• But in practice, where we do present information we believe is inaccurate, we do so in language that distances it. This is in accordance with the example above. We present the mainstream view in the simple declarative (e.g. Apollo program) and the fringe view surrounded by hedges such as "claimed", "allegedly" and "purportedly" (e.g. Moon landing conspiracy theory). Wikipedia always states the mainstream scientific or academic consensus as if it were correct. And it is right that we do so.—S Marshall T/C 15:36, 15 July 2011 (UTC)

Of course. That is how we give an overwhelmingly majority viewpoint its due weight. And, of course, figuring out due weight gets trickier when there is disagreement in the mainstream... at which point we present all viewpoints as "opinion". However... this has nothing to do with "Verifiability not Truth". We still exclude information that is not verifiable (no matter how many editors believe that information to be true). Blueboar (talk) 16:11, 15 July 2011 (UTC)

But that's not the problem. The problem is how to treat information that's verifiable but untrue.—S Marshall T/C 16:17, 15 July 2011 (UTC)

And that the current wording is widely interpreted as outlawing seeking accuracy. North8000 (talk) 16:30, 15 July 2011 (UTC)

"The current wording is widely interpreted as outlawing seeking accuracy." I disagree with this statement. I think most editors are aware that we are striving for accuracy. I would guess that among the public at large (i.e. those not intimately familiar with Wikipedia), the misunderstanding is the other way around: They believe that Wikipedia values Truth over Verifiability. The wording "Verifiability, not truth" is a wake up call for those people. Mlm42 (talk) 17:38, 15 July 2011 (UTC)

That's not always true, actually: We do not always distance ourselves from what we know to be errors. Sometimes what is verifiable is actually false, and is known to an individual editor to be false. I'm sure that I could find examples of editors who are writing within their field and who present the mainstream view, as it's known today, knowing full well that it was actually false in some particular point, and that the sources to demonstrate its falseness were still in the process of being published.

In such a situation, the editor would be entirely correct to write (or to leave in the article) "Last quarter, the company reported that it earned $____", even if he personally knows that revised figures will be announced tomorrow morning. "Truth" is not an excuse for committing original research. WhatamIdoing (talk) 16:52, 15 July 2011 (UTC)

For larger theories/concepts, disputed content or undisclosed (hence unverfiable) information I agree. However for smaller factoids that's not quite true. A more common scenario would be to correct quotes, typos or simple straight forward computation errors in secondary literature. The difference to your example is, that the correctness in such cases is verifiable by checking primary sources or performing a simple computation yourself and not original research either.--Kmhkmh (talk) 17:05, 15 July 2011 (UTC)

Yes, but off-topic: The complaint is about the tiny fraction of material that is either unverifiable-but-true (which we do not include) or verifiable-but-untrue (which we do include).

The trivial case that you describe (the material is not only verifiable-and-true, but also appropriate, encyclopedic, given due weight, on-topic and so forth) is not what's being discussed. WhatamIdoing (talk) 20:12, 15 July 2011 (UTC)

Blueboar/WhatamIdoing, are you seriously saying that in the part of Wikipedia that works, that for each fact that is put in the editor(s) found and tallied up what all of the sources said on it, found the majority view and then put that in? I assert that here is how it actually usually (= on non-contentious articles) happens: The editor(s) know the topic, which means they are carrying, in their heads, summarized information from dozens or hundreds of sources on the topic. They also know which sources are good, and which are crap ("crap" usually meaning not very useful rather than wrong) They either write the true material and then select and go to good source to cite it, or they select and go to a good source and use it to help write and cite the planned material. Odds are if you pick a sentence from today's FA and trace it back, that it will have come about in this way. North8000 (talk) 17:49, 15 July 2011 (UTC)

"are you seriously saying that in the part of Wikipedia that works, that for each fact that is put in the editor(s) found and tallied up what all of the sources said on it, found the majority view and then put that in?"... Yes, I would say that... on non-controversial topics there is rarely a need for this tally to be spelled out in a formal talk page debate... but it happens never the less. It occurs unconsciously, as the editors do research on their topic. When (as you say) editors "are carrying, in their heads, summarized information from dozens or hundreds of sources" they unconsciously become aware of which views are in the majority and what are in the minority. They mentally tally up what the different sources say, and assess them.

They even come to a conclusion as to which viewpoint they think is "true" (not necessarily the majority view). However, when it comes to writing an article, they must set aside their personal conclusion as to "truth". They must accurately (and neutrally) present all significant views, and give them due weight. They can not only cover the view they agree with (ie think "true"), they must also cover the views they disagree with (ie think are "untrue"). Blueboar (talk) 19:35, 15 July 2011 (UTC)

That's what WP:V in its present wording says they should do, yes. The question, I think, had more to do with how Wikipedia works in practice—where Apollo program says that the moon landing happens and mentions no other possibility; where global warming says very clearly that anthropogenic global warming does take place; where evolution says very clearly that organisms evolve; where taxonomy doesn't mention baraminology; where Elvis Presley says that Elvis is dead; or where Egyptian pyramids unaccountably fails to mention ancient astronauts. I could go on and on.

My point is that in practice, when the chips are down in the face of potential controversy, Wikipedians make judgments about what's true and what isn't, and write their articles accordingly. Policy ought to document practice.—S Marshall T/C 19:54, 15 July 2011 (UTC)

No: "When the chips are down in the face of potential controversy, Wikipedians make judgments about" what's DUE and what isn't. They don't decide what's True™ and what isn't. WhatamIdoing (talk) 20:06, 15 July 2011 (UTC)

The evidence says that contrary to Blueboar's earlier post, Wikipedians certainly don't cover all significant views, WhatamIdoing. The evidence is silent on what judgments Wikipedians make, but I'm personally quite sure that Wikipedians have views on what's true and what isn't. And I'm quite sure that these judgments influence what they write, too.—S Marshall T/C 20:27, 15 July 2011 (UTC)

Ah, I get it. You're pushing for this policy to be more descriptive, rather than being prescriptive. The thing is, on issues of principal such as this, we're actually not descriptive, intentionally so. People edit from a non-neutral perspective all the time as well, but it would be a huge mistake to say "well, we can bend our NPOV policy a bit, since no one wants to follow it about this point". Just because editors, even well-established editors, edit articles to add The Truth™, doesn't make it correct. That's actually the principal reason that I mentioned the Climate Change area, above.
— V = IR (Talk • Contribs) 22:23, 15 July 2011 (UTC)

Well, that's certainly a refreshingly different take on WP:PG, Ohms law.  :-) I'm afraid that it's original research, though; you're inventing a totally novel principle, not documented anywhere else and without any kind of consensus supporting it. In an encyclopaedia that's run according to the five pillars and WP:PG, policy documents practice.—S Marshall T/C 22:59, 15 July 2011 (UTC)

Sorry, but you're flat out wrong on this point. Mistaken, at least. The pillars are non-negotiable. Always have been. Most policy is descriptive, and that's actually one of our principals as well, but there is a prescriptive core to policy, and Verifiability is one of them. That doesn't mean that this policy can't change in minor ways, but it does mean that the essential meaning to it should never change. Go ahead and read 5P, or feel free to ask others about this point elsewhere.
— V = IR (Talk • Contribs) 23:51, 15 July 2011 (UTC)

You're agreeing with me. The pillars are non-negotiable. (They haven't "always been" non-negotiable, actually. The pillars didn't even exist until 2005. They were written after the policies. But they're effectively non-negotiable now.) What I said was that in an encyclopaedia that's run according to the pillars, policy (which is the layer beneath the pillars) is descriptive—including this policy. Do you understand now?—S Marshall T/C 00:26, 16 July 2011 (UTC)

Additionally, it's not what "an editor" does; it's what "all the editors collectively do". Although any given person, for any given point, might not extensively research something to see what "all" the sources say (although for material for which very few sources exist, that is not only possible but actually done, by myself and others), between us all, over the years, that is essentially what is done. We search for sources; we compare the sources; we report the balance of the sources.

None of which really has anything to do with WP:V. All these examples about whether we should mention ancient astronauts in Egyptian pyramids and the like are strictly matters for WP:NPOV. WP:V only concerns itself with whether the thing was said by a source that we could use to support material if that material is included. WP:V does not concern itself with whether that material should be mentioned in ___ article or in ___ manner. WhatamIdoing (talk) 20:06, 15 July 2011 (UTC)

Agreed. What we're talking about is the impact of two words which also have nothing to do with WP:V namely "not truth"  :-) North8000 (talk) 20:21, 15 July 2011 (UTC)

Largely, North8000, but I'd take issue with WhatamIdoing's last sentence too. In its present wording WP:V explicitly empowers editors to add sourced material, and explicitly doesn't care whether that material is true or false.—S Marshall T/C 20:29, 15 July 2011 (UTC)

Yes, but that's because of the limited role that this policy document encompasses. As WhatamIdoing mentioned before, this policy doesn't exist in a vacuum, you need to be aware of NOR, NPOV, IRS, etc... as well (the ideas behind them, at least).
— V = IR (Talk • Contribs) 22:06, 15 July 2011 (UTC)

... or we could just keep it simple and remove the words "not truth" from the policy. They're only a soundbite, Ohms law. The whole policy makes perfect sense without them.—S Marshall T/C 23:01, 15 July 2011 (UTC)

I disagree. The phrase "Verifiability, not truth" is the policy, for the most part. The rest of the page is there primarily to support that statement.
— V = IR (Talk • Contribs) 23:54, 15 July 2011 (UTC)

The verifiability policy is about the principle that edits must be verifiable, in that anything challenged or likely to be challenged should be supported by an inline citation to a reliable source. Properly understood, it's got nothing to do with truth at all.—S Marshall T/C 00:26, 16 July 2011 (UTC)

Ohms law, honestly, that's so off-the-wall that I don't even know how to respond. Except to say that I think that the other 99.9% of wp:ver shows that to be wrong. Sincerely, North8000 (talk) 00:45, 16 July 2011 (UTC)

More in reply to what Marshall said above (I can't find it in the wikitext...), the pillars have always existed, they just didnt' exist as Wikipedia:Five pillars until 2005. And I agree that policy is descriptive, but I don't see how you think that you agree with me when you're saying that the pillars are descriptive as well. You're not making any sense. Anyway, I agree with the section below. Stick → Dead horse. Time to move on.
— V = IR (Talk • Contribs) 01:24, 16 July 2011 (UTC)

Technically, "the pillars" haven't always existed. Various WP:Principles have always existed, but what those principles are, how many there are, and how to write them down has always been a matter of debate. "Five Pillars" happens to be the most popular embodiment of the community's principles, but it's not some sort of fundamental, non-negotiable, can't-exist-without-it page.

In an exactly analogous situation, "Verifiability" hasn't always existed—but the notion that you write down what you can support with a source, rather than what you personally believe is the truth, has long been the policy of the community. The fundamental, non-negotiable what's-in-the-sources-rather-than-what-you-believe idea is the only real policy. This page is merely the community's best effort at writing down the principle for the sake of people who don't already know the real policy. WhatamIdoing (talk) 01:53, 17 July 2011 (UTC)

back to basics

At this point, I will remind everyone of the primary reason for saying "The threshold for inclusion in Wikipedia is Verifiablity, not truth": If something is not verifiable, you should not add it... even if you are convinced it is 100% true. Blueboar (talk) 00:53, 16 July 2011 (UTC)

Agree. So let's stipulate that so that we don't posit debates that do not exist. North8000 (talk) 01:02, 16 July 2011 (UTC)

But this is part of the debate... this is why I strongly oppose the idea of removing the phrase. In this limited context, I would argue that the phrase is vital to the WP:V policy. Indeed it is the heart and soul of the policy. Perhaps we can better explain it... but it must remain in the policy. Blueboar (talk) 01:06, 16 July 2011 (UTC)

And despite the polls and rfc and months of discussion, there's no consensus to remove it. That horse is dead, let's move on. --Nuujinn (talk) 01:10, 16 July 2011 (UTC)

Also no consensus to keep it. I think we had some "middle ground" ideas of adding an explanatory sentence. But folks saying that such a middle ground compromise is too radical forces folks back to arguing their actual / further-from-the-middle positions. Sincerely, North8000 (talk) 01:36, 16 July 2011 (UTC)

Please, show me the policy allows a change on the basis of no consensus. I believe that we've been through enough discussion on changing "The threshold for inclusion in Wikipedia is Verifiablity, not truth" to establish that currently there is absolute no consensus for a change, and continuing to beat that dead horse is not productive. And I'm willing to work on wording in other sections, I'll make some suggestions tomorrow morning. What I see is a continuing effort to try to force a change where no consensus for a change exists. --Nuujinn (talk) 02:03, 16 July 2011 (UTC)

You misunderstood my comment. North8000 (talk) 03:25, 16 July 2011 (UTC)

Then please explain what you meant by "Also no consensus to keep it". --Nuujinn (talk) 04:23, 16 July 2011 (UTC)

First, answering your question, the meaning of consensus is hard to pin down, but it always means a substantial majority. Neither the "change it" nor the "keep it" viewpoint achieved that. Now on to the mis-understood (at best) part. The impasse is about changing that sentence. I brought up the middle ground ideas which are adding an explanatory sentence without changing the one in question. You, with strong words ("beating a dead horse") implied that I was arguing for the change that had the impasse. Sincerely, North8000 (talk) 12:38, 16 July 2011 (UTC)

Sorry, that was not directed at you specifically, just in general, I should have been more clear. I was emphasizing what Blueboar said, not responding to your comment on positing debates, and I overreacted to your comment about consensus, please accept my apologies. --Nuujinn (talk) 15:42, 16 July 2011 (UTC)

There's plenty of consensus to keep it. It's there, after all. Has been for years now, and the recent attempts to remove it have obviously raised significant opposition. It's not unanimous as long as you guys persist, but... I mean, there's more people willing to retain the current wording than there is to remove or change it, so in the long run I'm betting that the current wording will stay. By the way, I'm willing to listen to suggestions, or to see changes, which add additional explanation to the paragraph as well. Anything short of changing the phrase itself is probably acceptable, to me.
— V = IR (Talk • Contribs) 02:28, 16 July 2011 (UTC)

I agree, this discussion appears to be going nowhere. It was interesting to explore the relationships between "verifiability", "truth", and "accuracy"; but ultimately the campaign to change the opening sentence (and in particular, remove "not truth") isn't going to succeed without widespread support. At this point, it's obvious that there is not widespread support. It's time to move on. Mlm42 (talk) 02:33, 16 July 2011 (UTC)

Y'know, I actually think it's too early to decide to quit on this one. There have been some interesting suggestions raised which could be taken further and make for clearer wording / re-wording. I know we're all dead tired of this debate, but it seems a shame to collapse in a heap and leave it just as there's a possibility that we could work together on something that actually makes it better. Pesky (talk …stalk!) 06:27, 16 July 2011 (UTC)

In regard to the phrase "The threshold for inclusion in Wikipedia is Verifiablity, not truth", how many more months should we discuss it? I'm game for the rest of the policy, but seriously, how many more months for that particular phrase? --Nuujinn (talk) 06:33, 16 July 2011 (UTC)

Dunno, mate! How does "The threshold for inclusion in Wikipedia is Verifiable truth, not unverifiable truth" grab you? (Bearing in mind that the "verifiable truth" can be "it's true that X person made Y statement or has Z opinion). :o) Pesky (talk …stalk!) 06:55, 16 July 2011 (UTC)

Wikipedia:Drop the stick and back slowly away from the horse carcass. Mlm42 (talk) 07:11, 16 July 2011 (UTC)

There are no horse carcasses here, and no sticks.  There are, however, figures of speech (see new section below) including "horse carcass"; and two words, "not truth", ambiguously used as both a literalism and as a figure of speech.  Unscintillating (talk)

It's unfortunate and annoying that a lack of consensus means permanent stagnation, but I agree that this proposal is stuck; thousands of words and nobody's changed their position. However, axiomatically, consensus can change, and the lack of consensus over this particular wording is quite likely to change, over time. I'm sure it will be back.  :-)—S Marshall T/C 08:34, 16 July 2011 (UTC)

Perhaps. But thank you for dropping it... even if only "for now". Blueboar (talk) 12:25, 16 July 2011 (UTC)

The lead says with emphasis that not truth is acceptable if verifiable.  WP:Editing policy states, "...on Wikipedia a lack of information is better than misleading or false information—Wikipedia's reputation as a trusted encyclopedia depends on the information in articles being verifiable and reliable."  As long as we have editors believing that WP:V means that it is not only accepted but expected to include false but verifiable information in Wikipedia's voice, the problem has not gone away.  On the other side of the equation, editors don't get from reading this sentence that accurate information is not acceptable if it is not sourced.  Unscintillating (talk) 16:57, 16 July 2011 (UTC)

No... the emphasis is the reverse... to indicate that assertions of truth are not acceptable if not verifiable. We look to other policies (especially NPOV) to resolve "verifiable, but not truth" situationsBlueboar (talk) 17:25, 16 July 2011 (UTC)

"Not truth" is ambiguously both a literalism and a figure of speech

While this conversation has reached a lull, this seems to be a good time to step back and discuss why we haven't been able to resolve the issues.  One point is that we did not reach a consensus in the previous discussion about "not truth" being a figure of speech.  Perhaps the problem is that it is more accurate to identify the two words as ambiguous in having both a literal meaning and having a figure of speech meaning.  Unscintillating (talk) 16:57, 16 July 2011 (UTC)

I am not sure I understand your distinction. But do we really have to re-open the discussion yet again? The phrase says exactly what it means... It does not matter if you are 100% positive that something is true... if it is not verifiable you can't include it. It really is that simple. Blueboar (talk) 17:33, 16 July 2011 (UTC)

Argh! I was trying to drop this, but I literally can't let that pass. If it said that, Blueboar—if the wording was exactly what you just wrote—then I wouldn't have a problem with it at all! But that's not what it says.—S Marshall T/C 21:22, 16 July 2011 (UTC)

Huh? That's exactly what it says. Blueboar (talk) 21:29, 16 July 2011 (UTC)

What about adding that sentence as a follow on to the sentence that is there? So, we'd have The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. The fact that an editor or editors are 100% positive that something is true is irrelevent; if it is not verifiable you can't include it. for the first paragraph (slightly edited to make it more appropriate for the document). I think that it's simply restating what's already in the second sentence, but I wanted to offer something like this as a sort of compromise.
— V = IR (Talk • Contribs) 23:08, 16 July 2011 (UTC)

Maybe we should change it to "The threshold for inclusion in Wikipedia is Verifiability, not awesomeness".. but then people might complain that Wikipedia is anti-awesome? Seriously, this is basic English grammar; it means exactly what Blueboar said it means. We should really drop this before we all lose our minds. Mlm42 (talk) 21:34, 16 July 2011 (UTC)

Blueboar, what it says is that it doesn't matter what's true, it only matters what's verifiable. Which is a very different thing from what you just wrote. Did you seriously not understand where North8000 and Unscintillating and Hans Adler and me and, well, everyone else who doesn't like the current version were coming from?—S Marshall T/C 21:38, 16 July 2011 (UTC)

1. So one source of difficulty is that editors don't agree what the two words mean.  Unscintillating (talk) 21:39, 16 July 2011 (UTC)

Wishing the best for everyone's minds, #A suggestion, above. --Tryptofish (talk) 21:40, 16 July 2011 (UTC)

Structurally, the operative part of the sentence is "The threshold for inclusion in Wikipedia is verifiability", it is not affected by any listing of what it is not. You could list a hundred things that it is not, and it would not affect the operative clause. The effects of "not truth" do not come from the structural statement of the sentence, the come from the impressions that it leaves, and the mis-quotings that it encourages. I think that the intent and benefit of the "not truth" is to address one of the most common challenges of the policy, where someone essentially says "but it's true, so verifiability should not be required". I personally think that this benefit is small/irrelevant, because whoever says that will certainly and always lose, as wp:ver is clear, explicit, strongly worded, and a policy. Speaking only for myself, I think that the main damage from "not truth" is that it contributes to various widely accepted damaging mantras and misconceptions which pretend to be derived from policy but aren't. Some of these fall along the lines of "editors job isn't to apply intelligence, seek accuracy, select accurate sources, or do what editors do to create all of the successful WP articles, it it just to watch good articles spontaneously arise from tallying and repeating sources." And if there is a POV battle, that accuracy is not even allowed to enter into the conversation. North8000 (talk) 22:33, 16 July 2011 (UTC)

Argh... I think I am about to loose my mind. Stop over-thinking this. Most people clearly interpret the first sentence in a manner very similar to what Blueboar said above. If you parse individual words with various meanings, you can twist around the intent of nearly any sentence uttered in English discourse, which is what I see you doing here North. One major reason such arguments are unconvincing is that they're not grounded in any actual commentary. Can you provide some diffs where you think that someone is espousing the interpretation that you're offering?
— V = IR (Talk • Contribs) 23:03, 16 July 2011 (UTC)

Do you have any diffs to show that "most people clearly interpret the first sentence in a manner very similar to what Blueboar said above"?  Unscintillating (talk) 23:27, 16 July 2011 (UTC)

If most people interpreted it that way, then we wouldn't have had such a conflict on this page in the first place and we wouldn't have so many news articles and organizations accusing Wikipedia of advocating a policy of promoting false information. I would definitely argue that most people interpret not truth as meaning that Wikipedia has no problem with pushing false information, just so long as it is verifiable and that all information on Wikipedia must be thought of as false, because Wikipedia has no reason to try to make sure it is true. That's the interpretation I think the world at large takes from "not truth". Silverseren^C 01:21, 17 July 2011 (UTC)

I disagree: Conflict involving a couple of pedantic folks on a policy page does not prove anything about the normal person. See, e.g., the ArbCom case about MOSDATE: 99.99% of editors—a fraction that qualifies as "most" by any standard—were uninvolved and unaffected.

I haven't seen a single reputable media source that actually asserts Wikipedia advocates promoting falsehoods.

I think if most people misunderstood this page, there'd be a different person howling about it at RSN every day, and lengthy discussions about its true meaning every month. This isn't happening. Not one of the people alleging a misunderstanding has ever put forward even a single diff showing that anyone except themselves has actually misunderstood it. And, yes, we've asked for those diffs, for months now. At this point, absence of evidence of a real-article, real-editor problem looks to me a whole lot like evidence of absence of any real disputes involving this hypothetical misunderstanding. WhatamIdoing (talk) 02:02, 17 July 2011 (UTC)

2. Another source of difficulty is comments not grounded by diffs.  Unscintillating (talk) 23:27, 16 July 2011 (UTC)

I figured someone would try to be smart and ask that. It's rather difficult to prove a negative. I'm not sure how I can show that people have not used the interpretation that north (and yourself?) are putting forward... other than not pointing to anything. As for showing that the policy as written is used in the manner that myself and Blueboar, among others, are saying, I could pick any of thousands of talk page conversations where WP:V is mentioned. Is that really necessary? the policy is as it has been for years now, I'm simply asking for some basis in fact that it's not always interpreted as nearly everyone apparently thinks that it's interpreted as (other than an apparently tiny minority).
— V = IR (Talk • Contribs) 00:54, 17 July 2011 (UTC)

3. Another source of difficulty is no agreement that a problem exists.  Unscintillating (talk) 01:57, 17 July 2011 (UTC)

4. Another source of difficulty is that the Discussion page format makes archives of diffs inconvenient to locate.  Unscintillating (talk) 02:47, 17 July 2011 (UTC)

Regarding examples that support what I said, ( from the many that I have drawn the conclusion from) they are pervasive; there is even at least one in this current talk page. If this is just User:North8000/Page2#Useful method to take a whack at any thought then it would be a waste of time, But if some folks are genuinely open to see if they lend support to what I said, possible say something, the I would be happy to work on it. North8000 (talk) 14:32, 17 July 2011 (UTC)

OK I found one right in this talk page. It the full ramifications covered by: "Our job here is not to present 'truth' but to present what reliable sources say about the subject" The full scope of this sentence includes excluding all other considerations (such as accuracy, plausibility) in deciding what does and doesn't go into an article. North8000 (talk) 14:37, 17 July 2011 (UTC)

See though, that's kinda my point. That's not an example, it's your own interpretation of what others are saying, and that interpretation isn't what most others appear to be taking. It certainly doesn't seem to be my own interpretation... although, it's kinda hard to tell because what you're saying is kinda written in double-speak. I know that I'm open to listening to any suggestions, and I have been listening here; it seems as though you're unwilling to consider anything but removal of "not truth", which is never going to be accceptable to this group of participants.
— V = IR (Talk • Contribs) 15:27, 17 July 2011 (UTC)

While I can't say exactly what that one person had in mind, such sayings are pervasively applied to mean exactly that. On "you're unwilling to consider anything but removal of "not truth", you must have missed a whole lot of things that I've written, such as the latest below. North8000 (talk) 15:31, 17 July 2011 (UTC)

5. Another source of difficulty is resistance to consensus building.  Unscintillating (talk) 15:56, 17 July 2011 (UTC)

6. Another source of difficulty is that sufficiently advanced incompetence is indistinguishable from trolling. —chaos5023 (talk) 16:25, 17 July 2011 (UTC)

"Not truth" discussion

Right now on CNN. Count Iblis (talk) 01:56, 6 July 2011 (UTC)

Link? Unable to find it on Google news or CNN's own Wikipedia keyword search. Jclemens (talk) 06:51, 6 July 2011 (UTC)

I can't find it either. However, I'm a little late to this party and didn't know about this until I saw a thread on S marshal's talk page about this while responding to something else there. I can see both sides to this. Some of the old timers have probably noticed that almost every POV pusher who has come along has claimed to bear "the truth" and I have no doubt that many of them think we should just AGF and accept their "truth". The first sentence is a polite way of telling them that what they offer, even if it's the "truth" is not enough. It should be possible for any reader to click those little blue numbers and decide for themselves if it's the "truth".

On the other hand "accuracy" is important too. It's quite possible for something to be "verifiable" at one time and later found to be weapons grade bullshit. We shouldn't be telling people that it's not important and that's what the first sentence might imply to some.

I think there should be a way of making both points to newcomers and others. I don't know what that might be but one idea might be to remove "truth" from the first sentence of this policy and expand WP:VNT into a guideline. This way we can both emphasize that accuracy is important but still preserve the old but important "verifiability not truth" meme and have something that isn't "just an essay" to point newcomers and POV pushers to when they claim to bear the "truth". --Ron Ritzman (talk) 03:14, 8 July 2011 (UTC)

That's an interesting idea. I'm one of those who has been fighting against removing 'not truth' from the lead. However, I'm open to doing so if Wikipedia:Verifiability, not truth is made into a guideline, with the provision that it not be in turn attacked by those who have been campaigning for the removal of 'not truth' from the lead here. LK (talk) 05:46, 8 July 2011 (UTC)

The trouble with such compromises is that, as any proper logician or mathematician will tell you, a compromise between a wrong solution and a correct solution is another kind of wrong solution. My position is that we need to remove "not truth" from this policy without elevating VNT to a guideline. It's a sin against the basic purpose of an encyclopaedia to publish known error, so we need to tell the truth. To actually instruct editors that they need not tell the truth is simply the wrong approach.—S Marshall T/C 07:20, 8 July 2011 (UTC)

I think that, whatever we do, we have to make sure that we're covering all the points in a way which is easy to understand for the people who are most likely to misunderstand it. How we actually word these things can have a major impact on how readily they're followed. This means that, ideally, the absolute simplest way of being accurate needs to be worked on. The purpose of a lesson isn't to show how erudite the teacher is - it's to get the lesson across to everyone! Pesky (talk …stalk!) 07:36, 8 July 2011 (UTC)

Incidentally, it strikes me that we need to merge WP:VNT and WP:TRUE back where they belong, into WP:V. Or just delete them, if we think WP:V is the consensus version.—S Marshall T/C 11:22, 8 July 2011 (UTC)

We have had multiple discussions, and even formal RFCs on this issue in the last few months. Please, stop beating this particular dead horse. It is getting to the point of being disruptive. Blueboar (talk) 12:54, 8 July 2011 (UTC)

• Ron Ritzman's remark shows that there's still an appetite for discussing it among previously uninvolved editors, Blueboar. Please stop trying to dictate what other users discuss on this talk page. That kind of controlling behaviour is quite unacceptable.—S Marshall T/C 13:08, 8 July 2011 (UTC)

It's an important, current, pervasive problem. On the last specific proposal for changing the first sentence there was no consensus to change nor consensus to keep so it stayed with the status quo. Most likely if someone came up with a change which at least mitigated the problem without changing the first sentence it would pass. How 'bout adding a "radical" :-) statement like this just after the infamous sentence: "Verifiability helps achieve accuracy" North8000 (talk) 13:23, 8 July 2011 (UTC)

I am not trying to dictate anything... I am merely pointing out that the idea of dropping the "not truth" wording has already been discussed (repeatedly)... and that the idea has been rejected (repeatedly). To continue to push for it smacks of WP:IDIDN'THEARTHAT. If we are going to resolve this issue, we have to find another way to do it... something that can gain a clear consensus. Blueboar (talk) 13:28, 8 July 2011 (UTC)

Thanks for that clarification. On a separate note on the later part of your first sentence, If you are calling the lack of a consensus "rejected", the the idea of keeping the first sentence as-is has also been "rejected". North8000 (talk) 13:52, 8 July 2011 (UTC)

But I think that the reality is that I don't think that we're going to change the first sentence so we should fix the problem by other means. North8000 (talk) 14:03, 8 July 2011 (UTC)

• I don't think it's been "rejected" at all, Blueboar. I think, rather, that there is no consensus in favour of it—an important distinction because there is also no real consensus in favour of the current version. The reason the "not truth" phrase remains in the policy is not because there is a consensus to support it, but because on Wikipedia, a lack of consensus results in stagnation. In other words, SlimVirgin's wording from 2005 remains because of her first-mover advantage rather than because it really enjoys widespread support. It's unfortunate that sheer RFC-fatigue will maintain this unsatisfactory position for the time being, but there's no reason why this should prevent discussion on the talk page.—S Marshall T/C 14:53, 8 July 2011 (UTC)

What about linking Wikipedia:Verifiability, not truth in the three words? That way, the meaning can be fully explained elsewhere. Cambalachero (talk) 15:49, 8 July 2011 (UTC)

• I get the feeling that discussion in this talk has a tendency to go round and round. I really believe that not only will "not truth" be retained, but it should be retained. But the accuracy distinction is an excellent one. Lest it be lost, please see #A possible middle ground?, above. I think that might be the way to an acceptable fix. --Tryptofish (talk) 16:39, 8 July 2011 (UTC)

The discussion I was referring to was the one about the verdict in the Casey Anthony case. There was quite an interesting discussion on CNN about this with lawyers, a former judge and a few prosecutors. It had quite strong parallels with the "not truth" discussions we've had here. Someone in fact mentioned that in general judical cases are not about the truth, but some others disagreed about that. Count Iblis (talk) 16:55, 8 July 2011 (UTC)

I'm new to this discussion; maybe we need more 'new' voices in, to see if a consensus for anything can be achieved with some more and / or newer voices? I can see the frustration in here, on both sides. We all need to remember that each of us is actually trying to make things better; each of us is acting in good faith for something that we genuinely believe in. It's very hard when people get frustrated, impassioned, all-around fed-up, and all the rest! But - have faith, we're actually all attempting to work towards improvement of some kind.

I personally think that there's room for merges, room for change, and all the rest. I don't think that we necessarily need multiple pages to cover the subject - the fewer places we're directing people to, the better, on the whole. But we do need to keep discussions going - sometimes something new just turns up in a discussion and everyone says "Why didn't we think of this before?"; sometimes consensus changes simply because something's been presented in a slightly different, more clear, way; sometimes it's just new voices - but, whatever it is, if end up with improvement, it's a good thing. After all, we're (presumably) all working consistently towards trying to make things better? So - take a few deep breaths, have a beer / cup of tea / whatever, chill out a little, and recall, whenever we get frustrated, that we're actually all on the same team! [Pesky offers soothing hugs to whoever feels in need of same :o) ] Pesky (talk …stalk!) 05:59, 9 July 2011 (UTC)

The problem is that a lot of editors interpret verifiability, not truth as something like verifiability, regardless of accuracy or verifiability, not factuality. In my opinion, what really matters isn't so much that the misleading formulation is changed, but rather that we document unambiguously that this is not the intended meaning. So far, whenever I tried to explain this to one of those verifiability warriors, I was told that I am just plain wrong and don't understand Wikipedia. E.g., if the Register claims that A happened in Wikipedia, and the Wikipedia page history, in accordance with what respected members of our community remember, proves that the opposite is true, then we are still under an obligation to state as a fact that A happened even though we know it's false and libels someone. I am not making this up, this was a real dispute in which even an admin argued for libelling a former arbitrator because in his opinion verifiability in a purely technical sense trumps what we know reliably though not from a reliable source. It's a blessing that the Register's lie wasn't picked up by other reliable sources while we were repeating it without qualification, under the assumption that Wikipedia must be the most reliable and most direct source for such claims about Wikipedia, and that Sam Blacketer was not interested in suing the Wikimedia Foundation.

Regardless of any definite outcome, these discussions serve an important purpose. If they had happened a few years earlier, I would have had something to point to in previous instances in order to educate the fundamentalists that their interpretation is a radical minority interpretation. Better still, as more and more editors participate in this discussion and develop their own approaches to what verifiability, not truth means, general awareness of the problem increases and it becomes more and more likely that editors inclined towards fundamentalism are corrected at an early stage, and that in case of an escalation the number of editors who defend common sense and the integrity of the encyclopedia reaches critical mass more quickly than happened in the past.

That's not to say that an explanation in the policy itself wouldn't be a good thing, too. Hans Adler 07:12, 9 July 2011 (UTC)

I can fully support continued attempts to better explain what we mean by "verifiability, not truth". I strongly oppose the idea of removing "verifiability, not truth". Blueboar (talk) 13:46, 9 July 2011 (UTC)

I agree that it's an important concept and what I first suggested was only one possible approach to differentiating between "truth (accuracy)" and "truth (belief)". Another idea would be to change it to "verifiability not woo woo" but I don't think that will fly :) --Ron Ritzman (talk) 00:29, 10 July 2011 (UTC)

Thus, I think it would be a good approach to keep the existing sentence about "not truth", but to follow it with a new sentence to better explain what it actually means. --Tryptofish (talk) 17:06, 9 July 2011 (UTC)

Sounds like an OK compromise. Let's work on it. North8000 (talk) 17:19, 9 July 2011 (UTC)

• Well, I don't agree. What'll happen is that editors will quote "Verifiability, not truth" at each other and disregard the subsequent clarifying sentence. The best approach is to remove the words "not truth" from the policy.—S Marshall T/C 14:37, 11 July 2011 (UTC)

• I agree with you that the "not truth" in that first sentence should go. Probably the most damaging two words in all of Wikipedia. But it doesn't look like it's going to happen in the near future. So an explanatory sentence after that to reduce the harm that "not truth" does is a "half way" idea. I would argue that a small step forward is better than none at all. But you could also argue that the small step forward could reduce the impetus for the bigger step forward of removing "not truth". Sincerely, North8000 (talk) 14:51, 11 July 2011 (UTC)

• I'm not sure that it won't happen in the near future. I think before making changes, it would be best to watch what happens with the alarmingly mistitled "cults" case that Arbcom are about to accept, and be informed by the inevitable RFCs on that. The case is basically about misleading content on Wikipedia, with specific reference to our coverage of certain right-wing US politicians (be it allegedly highly laudatory, as with Kenneth Dickson, or allegedly highly pejorative, as with Rick Santorum). The point I'm coming to is that the infamous "not truth" clause is quite arguably one of the most enabling things for editors who wish to write misleading content. I hope that in the light of that case, there will be some well-considered edits made to our policies.—S Marshall T/C 15:02, 11 July 2011 (UTC)

• I'm afraid that such ArbCom cases on disputes on politics pages will make it even harder to get rid of "not truth". We have to consider how we got stuck with this "not truth" clause in the first place. Obviously, on those topics were "truth" is not easy to determine objectively and opinion as published in sources is the only thing that's verifiable, you tend to have the most disputes. These disputes influence the policies we have on Wikipedia. Then on other topics were truth is a more important factor to consider than simple direct verifiability (e.g. in case of scientific topics), you have far less disputes so the editors on these pages have had far little say on the shaping of the policies (the main influence here has been on dealing with fringe science, pseudoscience, POV pushing on politically charged issues like climate change denial, etc.)

As a result, the policies have evolved such that the normal editing practices on certain types of scientific articles are now in conflict with a strict interpretation of some policies. The best way to fix things, i.m.o. is to draft completely new texts for policies instead of trying to incrementally change things. The problem with the latter apprach is that the text of the policy pages have evolved to become optimally adapted to dealing with problems on politics pages, so any small change will make things worse from the perspective of the regulars here. Count Iblis (talk) 15:29, 11 July 2011 (UTC)

Another angle on this is that a source can be judged unreliable for particular items of information based on whether editors think it is true. Such an editorial judgement of unreliability negates verifiability and due weight and justifies suppression of the source and the info it contains. See currently ongoing discussion about this here. Wtmitchell (talk) (earlier Boracay Bill) 20:47, 11 July 2011 (UTC)

Since the world is full of wp:"reliable" sources that are sometimes real-world-unreliable and wrong on the topic at hand, it is a normal practice (on non-contentious articles) for editors to try to figure out which ones are right and use them. This is only heresy per the "not accuracy("truth")" train wreck religion; exercising editorial judgment to not use wrong sources is how good articles are written and does not violate actual policy. (unless there is a pov question in which case wp:npov says that the wrong ones also stay in) But on contentious articles this whole process breaks down, and the poor wording of the first sentence is one of the contributors to this problem. North8000 (talk) 21:08, 11 July 2011 (UTC)

As opposed to the train wrecks we will get by removing "not truth"--from people who will argue endless that we cannot use a reliable source because it is not true, and that we can use an unreliable source because it is true. Remember, whatever we do here has to work everywhere on WP. --Nuujinn (talk) 21:25, 11 July 2011 (UTC)

The words "train wreck" are hyperbole and the words "we can use an unreliable source because it is true" are a straw man.  As a logical fallacy, the defeat of the straw man does not actually lead to the conclusion stated, which basicly is saying that POV truth-pushers can only be controlled at Wikipedia with the force of a particular two-word figure of speech.  Unscintillating (talk) 01:50, 12 July 2011 (UTC)

• Whereas if we say "verifiability, not truth" then we get the situation we have at the moment, where policy says that whatever can be sourced can be said. The trouble is that the sources we call "reliable" get it wrong all the time, and some of them don't get it wrong by accident. I mean, the BBC are one of the world's more reliable sources, but even they faked footage of child labour in a hatchet-job "exposé" of fashion retailer Primark. And I dread to think how much of Wikipedia is sourced to Fox News.—S Marshall T/C 21:50, 11 July 2011 (UTC)

Nuujinn, unless you are saying that something being sourced is a sufficient condition to force inclusion, then "we cannot use a reliable source because it is not true" is something that they can do right now, and rightly so. And, for challenged material, getting rid of the problematic words would not change that wp:ver says that wp:rs'ing is required to keep in challenged material. North8000 (talk) 23:59, 11 July 2011 (UTC)

No, I am saying that I think deleting use of "not truth" will create more problems than it will solve. I will also say that this horse has died, sunk into the earth, become fossilized, put on display in a natural science museum where it has been around long enough to have become dusty. By continuing this line of discussion for months when it is absolutely clear that there is no consensus for the change, some editors have, in my opinion, become disruptive to the discussions here. --Nuujinn (talk) 00:13, 12 July 2011 (UTC)

Do you have an example of where the two words avoided a problem?  Sorry, but we do not have a consensus to not change this wording—hyperbole is not a substitute for the force of reason.  Unscintillating (talk) 02:01, 12 July 2011 (UTC)

Just to keep terminology straight, since actually reliable sources often don't qualify as wp:reliable sources, and wp:reliable sources are often unreliable or wrong, when referring to a source defined as meeting wp:rs criteria, we should clarify that it is a "wp:reliable source", a distinction from "reliable source". North8000 (talk) 23:58, 11 July 2011 (UTC)

Thank you for that perspective.  Unscintillating (talk) 02:01, 12 July 2011 (UTC)

Well, it has been a long time since I've contributed to this page or the WP:RS/N so maybe Blueboar can correct me if I'm wrong–a "wp:reliable source" is not something that is inherent to a particular source, the NY Times is not inherently considered a wp:RS for anything stated in it no matter what, policy/guidelines are clear that a RS is only reliable for facts that fall under what is considered to be the source's primary area of expertise. You can't take Scientific American or Nature magazines if they give the wrong birthday for George Washington and claim now on the article that there is dispute on his birthday (which there kinda is because of the change in Gregorian/Julian calenders in Britian after he was born). Fox news is OFTEN taken to the RS/N for whether it is a RS in a certain case, and often it isnt an RS. The basic fundamental flaw with "verifiable, not truth" is that to be an RS you first have to be CORRECT. If you have one source that is an RS normally, and it has been shown by another RS to be false, then it is FALSE, you dont get to write in the article that there is a dispute or "one source says one thing another says another", it is simply false. Of course sometimes things arent so cut-and-dry and despite NOR sometimes we have to make judgement calls on these things. You may not want to remove "verifiable, not truth" but in the end TRUTH matters and since policy itself is clear that policy is not prescriptive but instead is descriptive ie- "policy describes best it can our current method of dealing with certain situations that have occurred already, use this as a guide to resolving future disputes, use with caution as specifics may vary and consensus may be different". So the question I pose is–is how is "verifiable, not truth" used by the Community in every day editing, once we resolve how the Community actually acts we can then describe it better in a policy or guideline. You (as in any editor) may not like how it is used, but policy is not about forcing Wikipedia into what you would prefer, it is about describing what is already done by the Community.Camelbinky (talk) 06:10, 12 July 2011 (UTC)

This is a good point - sometimes even the most "reliable" of sources can contain something as simple as a typo (on dates of birth, for instance) which escapes the sub-editor before it goes into print. Reliable sources can (and do) make honest mistakes - so we do need, from time to time, to make that judgment call on whether or not to include that stuff. For the most part, it's obvious, but sometimes it's less obvious. I think we may have kinda wandered off the main route here, though - I'm still mulling over how to put the idea simply that, no matter how true something is, it does have to have been scholarly-accepted-as-true (or at least as a possible alternative truth) before we can include it in wp. It's closely linked to the WP:OR thing, too. True and verifiable obviously has to take priority over published but inaccurate and also over "newly true" but not yet published in an accurate and reliable source. The second two kinds of info aren't acceptable for inclusion - only the first is. How about verifiably true? Does that get any closer? Pesky (talk …stalk!) 08:11, 12 July 2011 (UTC)

• This is a policy, so say simple things in accessible constructions. Begin with the basic idea, not the difficult cases. My ideal version of this policy might begin: "Before any kind of material may be included in Wikipedia, it must be verifiable. The best way to verify things is by an inline citation to a reliable source. In practice, not everything has to be verified, but if anyone challenges something, or if it seems likely to be challenged, please provide an inline citation to a reliable source..."

  After we've outlined the basic principles, we can go on to discuss the hard cases, such as how to deal with material that's verified but not true, or indeed true but not verified. But that first paragraph must be simple and clear.—S Marshall T/C 09:07, 12 July 2011 (UTC)

Sounds good. North8000 (talk) 10:57, 12 July 2011 (UTC)

Let me apologize in advance for being both pedantic and long winded, I recognize these are some of my many faults.

"The threshold for inclusion in Wikipedia is verifiability, not truth" is simple, and to my way of thinking, is the basic principle we need to articulate. Unscintillating, I believe that, generally speaking, consensus is required to make a change, but not to leave things as they are, so in regard to "....we do not have a consensus to not change this wording", I don't think we have to have consensus to leave "The threshold for inclusion in Wikipedia is verifiability, not truth" as it is. In terms of where it matters having "not truth" in the policy, I could say answer that it is an important phrase for any area where people disagree about what the truth is, which (IME) includes pretty much any article on the Balkans, politics, controversies (esp. if the sources are limited to eastern european or western asian news sources), fringe "science", or any of the other really messy areas where The Truth is articulated by believers. And when you dig into the topic, Truth is not a simple matter, see esp. Feynman's quote on that page.

But the real reason I like the phrase so much is it provides clear direction as to the mechanism we use to resolve disagreements about what to include and what to reject. Truth is contextual, not absolute. The sky is blue, but the sky is not blue. We will sometimes disagree about what is true in a given article, and the only way I can see to resolve such disagreements is to turn to reliable sources and use them to verify what is true in the context of that article, which is enough work as it is. "verifiability, not truth" gives us a simple, quick way to explain that one of the most important things we do here is not determine what is true, but what we can verify. It is an excellent phrase to slow down the tigers amoung us, the little black dress that is appropriate in all venues in WP when we are trying to determine what to add and what to remove. And since we have to use reliable sources to verify what we would call the truth in every case where there is a disagreement, we are really, in practice, talking about verifiability rather than truth even when we're trying to talk about what is true.

Now, I certainly endorse the notion that we can improve the wording and how we explain these concepts. I do sympathize with those who do not like the phrase, as there are many policies and guidelines containing phrases I'm not comfortable with. I'm also sure that many editors will disagree with what I've said here, and that's fine, as well-meaning productive editors often disagree. But if one examines the discussions above and in the archives, it seems pretty clear there's simply not consensus to remove the phrase. And many of us are quite honestly tired of talking about this--this issue has completely dominated this page for months now and is, I believe, interfering with our ability to get things done here. My suggestion is to let it go, and let's work on explaining it better. --Nuujinn (talk) 11:25, 12 July 2011 (UTC)

I'm sorry that you're weary of this discussion, Nuujinn, but I'd remark that nobody's changing the policy and nobody's forcing you to participate in the conversation. There is no disruption taking place here, or at least, not on the pro-change group's side. The reason the issue has completely dominated this page for months is because it's important, and because at a fundamental level there is genuinely no consensus about it.—S Marshall T/C 11:42, 12 July 2011 (UTC)

"...a fundamental level there is genuinely no consensus about it, yes, exactly, which is why we should drop it and move on. There is no sign of consensus on the horizon, the horse is dead. But I've said enough. --Nuujinn (talk) 13:50, 12 July 2011 (UTC)

S Marshall, IMHO you are absolutely correct in your comments about those two words, but as a compromise, what if, instead, we worked on an immediately-following sentence that mitigates the problems caused by those two words? North8000 (talk) 14:10, 12 July 2011 (UTC)

• My position would remain that we need to remove "not truth" (and also change "threshold") from that first sentence, but I wouldn't try to stop you.—S Marshall T/C 16:11, 12 July 2011 (UTC)

Requiring information to be "verifiably true" will be understood as meaning that editors get to remove anything that they believe to be untrue. We've already got this problem at hundreds of articles on controversial subjects. To give one minor example, I do not want to people with Multiple chemical sensitivity to feel any more empowered than they already do to delete the well-sourced material about the disputes over the condition. We already have enough problems with people re-writing the article to make it sound like the problem is definitely caused by nasty synthetic chemicals (or strictly a hoax, depending on the POV). Telling them that Wikipedia may not include information unless it's agreed to be True™ would result in a blank page at that article (and many others).

Now imagine what would happen at Climate change or Homeopathy. Well-sourced, well-written articles often include information that someone believes to be verifiable, but NOT true. WhatamIdoing (talk) 17:08, 12 July 2011 (UTC)

I don't think that the addition would need to be structural/operative, just as the problematic "not truth" wording is not structural/operative. It could be something as simple as adding "This does not intend the discourage seeking accuracy; verifiability helps achieve accuracy" North8000 (talk) 18:27, 12 July 2011 (UTC)

Is this a Brainwave? "Not only true, but verifiably true"

(Apart from being an almost arbitrary break to help with readability / editability) ... Neither verifiability nor truth, on its own, is what we want. Something true but not verifiable, or 'verifiable' (as in, someone can find an apparently reliable source for it) but not true, is not what we want.

What we're looking for is "Not only true, but verifiably true". I think (but then I'm biased, as I thought of it) that that covers everything we need for a short catchphrase. I also know that loads of people shy away from any kind of wording-change in horror / terror / whatever, just because it's that dreaded thing, change. We do need a nice short catchphrase, and we do need it to cover as far as possible the concept of true and verifiably so, and we do need something that won;t just let people think they can wantonly delete sourced material, and so on. I think getting the whole thing in one catchphrase is totally impossible (hence why we have a whole page, etc.) Bu t I think my idea has merit, as an evolved catchphrase over 'Verifiability, not truth.' :o) What do you guys think? I;m trying to be constructive and look for a solution here. Pesky (talk …stalk!) 09:56, 13 July 2011 (UTC)

Adding a bit - 'truth' changes; we all know it changes, as more research is done (in scientific fields) for instance, our concept of 'what is true' evolves. I think we need to get across this idea to people, so we're not seen as belittling anyone who knows 'the next truth' (one which has been researched but not yet published). Experiments can have unusual outcomes, so if an experiment with the 'wrong' outcome is published, it doesn't necessarily mean that its a 'new truth'. We could possibly, further down the page, cover something like that; there needs to be more than one experiment, by more than one set of people, ideally, for something to be accepted as a new truth. I can think of two examples straight away: schooldays, chemistry lab, making hydrogen. What's supposed to happen is that hydrogen collects in the test tube. What actually happens was that nothing happens, the chemistry teacher pokes the setup, the bung flies out and hits the chemistry teacher hard on the nose. Second: in a home trial on a large clan of cats, eight out of ten cats demonstrably actually preferred toffee-flavoured Angel Delight. Both truths - but not enoughly-true to take precedent over the more commonly accepted and expected truth! Pesky (talk …stalk!) 10:11, 13 July 2011 (UTC)

(edit conflict)No, we don't care if anyhing is true or not. We present what [WP:RS|reliable sources]] propose as what they believe is the truth, and if they don't agree, we present different opinions. We don't do any research to find the truth (which would be WP:OR, we don't exclude sources which we believe are not the truth, even if they are "reliable". Claiming that articles should present what is true will only lead to people excluding what they believe is not true, no matter how many reliable sources there are. (Note: obvious errors, typo's, outdated info from reliable sources should of course be removed: that's not what this discussion is about). If some reliable sources claim that diisease X is caused by pollution, and other RS claim that it is caused by genetics, then both views should be included in the article: an editor who removes one of the two opinions because the other is "the truth" should be reverted: as far as we know, neither may be the truth, one of them at least is probably false, but both are verifiable, and that's all that counts. Fram (talk) 10:16, 13 July 2011 (UTC)

Yes, that's true. Of course disease X may be caused by something totally different, and just need to be triggered, but we may not have found that out yet! Or it may be caused by both things .... But we do need some kind of verifiability for both 'versions of the truth'. And we need to teach people that we teach the controversy, if there is one ... Pesky (talk …stalk!) 10:42, 13 July 2011 (UTC)

Respectfully, someone saying on this very page "we don't care if anything is true or not" is proof of the damage that those two words have done. North8000 (talk) 10:55, 13 July 2011 (UTC)

No, it is not because of those two words that I don't care if anything is true or not, it is because I don't care about that, that I support the inclusion of those two words. Finding truth in primary sources is the job of secondary sources: we are a tertiary source, we compile what the secondary sources have found. Fram (talk) 11:04, 13 July 2011 (UTC)

Well said. I think what we're aiming for is to accurately reflect what reliable sources say is "true", in the context of the subject of the article and the sources. Presenting the Truth isn't our responsibility, and trying to do that just causes headaches. --Nuujinn (talk) 11:13, 13 July 2011 (UTC)

Terms and situations First, using the word "truth" instead of "accuracy" is a straw man, because one common meaning of the word "truth" is dogmas and opinions.

Next we must realize that there are three cases regarding "which is correct":

1. The metrics of a correct answer are agreed upon, but the answer is not agreed upon or known: E.G. Where did Emelia Erhart's final flight end? Another way to say this is that if full information were available, all reasonable parties would certainly agree. The leading edge of real science falls into this category.
2. The metrics of a correct answer are agreed on, (all would agree what "land a man on the moon" and "Win the Superbowl" mean) and the answer is overwhelmingly considered to be known. E.G. Did the US land a man on the moon? "Who won the 2011 Super Bowl?" Any alternative viewpoint in these situations is fringe.
3. The metrics of a correct answer are not agreed upon (how do you define "good" and "bad") nor is the answer. Example: Is Obama a good or bad president? In reality, "accuracy" is irrelevant/moot here, because both the metrics and the result are matters of opinion. North8000 (talk) 11:16, 13 July 2011 (UTC)

(ec)Also, I like North8000's suggestion regarding adding something like "This does not intend the discourage seeking accuracy; verifiability helps achieve accuracy", the wording is a bit awkward, but I think that's a good direction to go in. --Nuujinn (talk) 11:22, 13 July 2011 (UTC)

There are more cases, such as what context the answer has meaning or value. Examples included assumptions of a frictionless surface for experiments or situations in which one Truth (the earth is not a true sphere) conflict with another (the earth for most practical purposes is a sphere) and another (the earth is a chunk of stuff who's description is only roughly approximated by geometry). This is the essence of Pragmatism. Truth is a sticky topic. --Nuujinn (talk) 11:31, 13 July 2011 (UTC)

Agree on all counts. That's why my idea doesn't say "only put in accurate stuff", instead it just essentially says "we're not against seeking accuracy" North8000 (talk) 11:34, 13 July 2011 (UTC)

It seems that we're actually all agreeing on this, which is a good step forwards! "This does not intend to discourage seeking accuracy; verifiability helps achieve accuracy" is an excellent addition. The other stuff (the "nearly-true" or true for all useful purposes at this level comes straight from Wittgenstein's Ladder). Pesky (talk …stalk!) 11:56, 13 July 2011 (UTC)

fact vs opinion

To me, the debate over the "Verifiability, not Truth" clause centers on a simple question: When should we present something said by a source as being accepted fact (ie true), and when should we present it as opinion.

Wikipedia can not rely on the personal views of its editors (no matter how strongly held or how well reasoned) to answer this question. An editor may be biased (POV), or basing his/her reasoning on Original Research (OR). No, the views of Wikipedia editors as to whether the something is fact or not must be set aside. Instead, we rely on examining the broad corpus of reliable sources that discuss the topic (ie we see what lots of reliable sources say).

If the broad corpus of reliable sources essentially agree that something is fact, then Wikipedia should present that something as being fact... and contrary views should be presented as being minority opinions. If the broad corpus of reliable source do not agree on whether something is fact, then we must present both the majority and minority views as being "opinions". Blueboar (talk) 12:50, 13 July 2011 (UTC)

This is where the 'due weight' aspect comes in, really. If 70% of the sources take one view, then the article should be 70% weighted towards that point of view. The real problem (as always) occurs when we have passionate people on both sides of what is supposed to be an argument in the "debating" meaning of the word, but so frequently degenerates into an argument in the "flaming row" meaning of the word! Whenever people have really strong feelings that the other side of the argument cannot by any reasonable, conceivable means be "true", then we get the flaming row happening. And Wikipedia editors are kinda passionate people - we're driven by our feelings. We're certainly not driven by any cash incentive! So we're never going to be able to get rid of passion (nor would we want to). It's incredibly hard to get across the idea of either NPOV or due weight to people who cannot believe that anything said by the other side can be true or even reasonable. So ... how do we really get across the idea that editors should put their own feelings, passions and beliefs aside, and let the other side have it's own due weight in an article - without wanting to haul off and hit each other repeatedly over the head with spanners? Pesky (talk …stalk!) 13:01, 13 July 2011 (UTC)

• Blueboar, I think you're heading in the right direction with this. Policy should document what we actually do in controversial cases. Take examples like young earth creationism or global warming. What we do is to explain that there are various viewpoints, but then we clearly and unambiguously side with the scientific consensus—because in practice, what's actually true trumps the absurd fictions that some crackpot has persuaded a technically reliable source to publish. And I emphatically endorse our current approach. The only thing that's a problem is that our policy says it's not the approach we're supposed to have.—S Marshall T/C 13:15, 13 July 2011 (UTC)

I agree with almost everything you say S Marshall... but I disagree with your last sentence... I think the current policy (especially when read in conjunction with our other polices) states very well what we "actually do". We really don't care whether you (an editor) are convinced that something is true or not true. What we do care about is what the sources (all the sources, not just the ones that agree with your personal view) say, and we care about presenting what those sources say with neutrality and due weight. That's the key to understanding "Verifiability, not Truth". Blueboar (talk) 13:23, 13 July 2011 (UTC)

Well, respectfully, that's the exact opposite of what "verifiability not truth" says. In fact, young earth creationism and global warming represent the triumph of the scientific consensus over other published sources. Typically, we excuse this by saying that the peer-reviewed academic ones are the "more reliable" sources. But what that actually means, in the real world, is that academic sources are more reliable because they're more likely to be... the truth. Do you see how that directly contradicts "verifiability, not truth"? In fact, it's closer to, "truth over verifiability."—S Marshall T/C 13:29, 13 July 2011 (UTC)

We side with the scientific consensus because these are the more reliable sources. The chances that they present the truth or are closer to the truth than the fringe scores are of course considerable, but in the end, we prefer not to discuss the contents of such articles on the basis of truth, but on the basis of verifiability and reliability of sources. This may seem like semantics, but keeps things much more neutral and less heated. Discussing whether X is true or not true are pointless; discussing which sources have the best reputation generally, the most peer review, the most citations, the highest professional standards, ... may still get heated, but is in the end fundamentally different. Fram (talk) 13:38, 13 July 2011 (UTC)

It seems to me that what you saying, Fram, is that the good reputations, the most citations, the high professional standards, and the scientific consensus are routes to ... the truth. And we prefer them because they're more likely to be true. Isn't that right?—S Marshall T/C 13:41, 13 July 2011 (UTC)

No... we prefer them (and state what they say as fact) because they represent an extreme majority viewpoint. You may think what they say is "true"... but we don't care what you think. Blueboar (talk) 13:47, 13 July 2011 (UTC)

(edit conflict, respond to S Marshall's statement) In non-contentious articles, it works. Often editors decide which source is in error, which one is is correct, put in the material from the correct source and use it for the cite. Or decide on the correct statement, and then pick a source that supports it. This is not the heresy that some would claim, this is, in fact, in a major way, how successful articles are written. [IP redacted] User:Fred Bauder Talk 17:54, 13 July 2011 (UTC)

No, Blueboar, Wikipedia doesn't follow the majority viewpoint. See, for example, list of common misconceptions which by its very existence is a counterexample to that principle.—S Marshall T/C 13:55, 13 July 2011 (UTC)

We do so follow the majority viewpoint. It's just that the only viewpoints that "count" in our calculation of the majority are the viewpoints expressed in reliable sources, not the viewpoints held by undereducated people.

Also, that something is a "common" misconception is not the same as saying that 50.1% of humans believe the error. A majority of people don't believe most of those misconceptions. WhatamIdoing (talk) 15:24, 13 July 2011 (UTC)

Yes, WhatamIdoing, what I've been saying all the way through this is that we do pick and choose which sources to prefer. We decide which are reliable. This is not a democratic process, not a source-counting exercise, but a matter of editorial judgment in which we decide what we believe and what we don't believe. Which is, at its heart, the quest for accuracy, reliability, and other synonyms for "truth". And that's why it's perverse to pretend that verifiability isn't about truth. It is.—S Marshall T/C 15:31, 13 July 2011 (UTC)

(ec)What Blueboar said. We don't decide what is truth or what not, we set up rules for verifiability and reliability, and we trust the sources that meet our standards to get it right most often. But if they turn out to be wrong and the fringe, ormore likely new research is right instead, so be it. We will then adjust our articles to represent the new scientific or mainstream consensus: but we will not be shamed or humbled, we will not have failed, since it is not our function to present truth, nor to go searching for it: we are, again, not a primary or a secondary source, but a tertiary one. We need to compile, present, summarize, reflect the sources in an accurate, truthful, neutral manner, but we don't pick and choose to present one truth or the truth. Fram (talk) 13:59, 13 July 2011 (UTC)

• I'm deeply perplexed by this answer, Fram, because I've literally just linked to two high-profile examples of articles where we clearly present two different views from two different sets of sources, and then clearly say that the scientific consensus version is correct. What's this, if it isn't picking and choosing to present one truth or the truth? Shall I link to other example articles where we do exactly this? I certainly can...—S Marshall T/C 14:17, 13 July 2011 (UTC)

That's a different issue. Fram's reply is excellent, I wish I'd said that! Dougweller (talk) 14:33, 13 July 2011 (UTC)

(ec) Re: List of common misconceptions... Common belief is not the same as majority view in reliable sources... also, I will note that the editors of this list are fairly strict about requiring citation to reliable sources that actually say the items listed are misconceptions (and common). You can't just add something because you think it is a misconception (ie the article does not allow assertions of "truth" for inclusion). Blueboar (talk) 16:37, 13 July 2011 (UTC)

I don't quite agree with Fram's comment as (an absolute) general guideline and a part of the problem here is, that a general rule doesn't really cover all the different scenarios in WP that well. Truth and verifiability have slightly different meaning in different areas/fields and there is also a difference whether we are talking about sourcing marginal factoids or theories/concepts/ideas. In particular with regard for the former we do more than just trusting reputable sources to get it right, but (ideally) we actually verify them independently and if necessary correct them. If for instance a reputable source makes an obvious straight forward calculation error (at least obvious to domain experts and for non scientific fields you could think of acitation or copy errors as a similar scenario), then we correct it or at least note the error, but we're definitely not going to copy the reputable source blindly.--Kmhkmh (talk) 16:30, 13 July 2011 (UTC)

Conflicting 'truths'

I work mainly within the field of archaeology and history. In those fields 'truth' can be extremely elusive, perhaps impossible to determine depending on how you define 'truth'. And there are other fields in which there are conflcting 'truths'. Our job here is not to present 'truth' but to present what reliable sources say about the subject, because that's all we can do when reliable academic source are in conflict (note I'm sticking with academic sources here to make my point). Dougweller (talk) 14:37, 13 July 2011 (UTC)

For those areas, where it is elusive, I think that that simply means that it is not known, not that it does not exist. North8000 (talk) 15:10, 13 July 2011 (UTC)

OK here's an example Three ship experts are writing a WP article about the USS Flutaflammer. The need to put in the length. They (without making a life out of it) can find only two sources that say the length. Both generally reliable, except one made a mistake. One said 735 feet long, which all of the editors agree is plausible. The other said 2,735 feet which all of the editors agree is clearly implausible. The way it really happens is that they decide to say it's 735 feet, and cite the source that said that. Is somebody here saying that they should not have done that? North8000 (talk) 15:22, 13 July 2011 (UTC)

I agree. Some time ago I argued with an editor about an even more problematic issue; the only source making a statment was wrong. It was about a series expension of a function, and the issue was a typo in the source. That was resolved on the basis of "truth" (and perhaps we also violated the NOR/Synth rule). If it's easy to show that a source is wrong, why have wrong information on Wikipedia? If there is a genuine dispute about this issue that is not based on Wiki-fundamentalism about applying the rules here, than that's a different matter. Count Iblis (talk) 15:45, 13 July 2011 (UTC)

I agree as well (see also my posting one further up).--Kmhkmh (talk) 16:44, 13 July 2011 (UTC)

I wouldn't object to the resolution in that example... however, what the editors in question are doing is applying a combination of WP:RS, and WP:CONSENSUS. The editors agree that on this specific fact, one source seems more reliable than the other. We allow for this (if need be, invoking WP:IAR to justify it). But... consensus can change... if a group of editors comes along and challenges the previous consensus, arguing that 2735 feet is actually plausible. At this point the article should be changed... so that it presents the debate... mentioning both lengths and attributing them (as opinion) to their authors. And, as further reliable sources are found (or written), the article would have to be changed yet again, to account for these new sources. Blueboar (talk) 16:03, 13 July 2011 (UTC)

OK, now for the same situation except with a contentious article. A POV warrior (who knows that the 2735 is probably wrong, but where having it in there serves his purpose) enters the scene. He tells the editors that the process of trying to to determine which is right is forbidden in Wikipedia, and so the wrong number also stays in the article. This type of situation is VERY common.North8000 (talk) 17:06, 13 July 2011 (UTC)

• Actually I think it's simpler than that. Blueboar is right to say we decide what's more reliable. Of course, by "more reliable" what we mean is "more likely to be true".—S Marshall T/C 17:46, 13 July 2011 (UTC)

  That's a good example, but I find it to be overly contrived for essentially the same reasons that Blueboar gave here. Just because we need to remain neutral doesn't mean that we have to be irrational as well. Actually, I'd argue the exact opposite: since we're required to be neutral, we are required to make more rational decisions about the conclusions reached by reading sources. We should all strive to be collegial as well though, meaning that we should at least listen to what the "POV warrior" has to say. In the example given, a reply such as "most of us agree that a ship of this type being 2,735 feet long would be implausible because <such and such>, which makes <the second source> unreliable here in terms of the length of the ship." The "POV warrior" is probably arguing that <the second source> is reliable for something else, and a third (fourth?) editor is reverting them citing the unreliability of that source in terms of the ship's length. In my experience, the vast majority of "POV Warrior" accusations arise from that sort of situation (the others are, of course, consperacy theorists and fringe theorists, but those situations are less... pernicious). Listening to what the "POV warrior" has to say, which should come down to something like "I realize that <the second source> contains a typo for the ships length, but it's correct about <some other subject> and I'm being told that I can't use it for that because of the length issue", should make it clear that the solution is that <the second source> is perfectly useable for everything other than the length issue.
  — V = IR (Talk • Contribs) 18:40, 13 July 2011 (UTC)

Under my example, the POV warrior simply wants the wrong information in there, and can use the "not truth" mantra to say that policy supports keeping it in. North8000 (talk) 18:51, 13 July 2011 (UTC)

Unrealistic... people don't become POV warriors for no reason. Chances are, if someone feels strongly enough about the length of the ship to POV war over it, we overlooked (or chose to ignore) an important aspect of the length debate. In which case, we probably should have been more neutral and mentioned both numbers in the first place. At minimum, I would ask the POV warrior to better explain why he/she supports to 2735 number.

And let's not forget the third option... deciding to omit both figures (ie deciding that the length of the ship is really an unimportant factoid, one that is distracting us from writing a quality article). Blueboar (talk) 19:10, 13 July 2011 (UTC)

(added later) In a typical POV war, they are not pushing the ship length per se, they want it in there for impression only on a different item which they are warring. (e.g. that that Navy is mighty, or that that ship builder builds excessively large ships) ) North8000 (talk) 20:21, 13 July 2011 (UTC)

This was addressed by myself and Blueboar: "Chances are, if someone feels strongly enough about the length of the ship to POV war over it, we overlooked (or chose to ignore) an important aspect of the length debate."
— V = IR (Talk • Contribs) 20:30, 13 July 2011 (UTC)

And POV warriors don't stop at one issue, they almost always have an overall agenda that reveals itself pretty strongly, and at some point they get blocked or just ignored. We will never be able to write a policy that prevents editors from causing disruption in the first case, we have to deal with those kinds of issues with policies that help us shut down the disruptive behaviour. --Nuujinn (talk) 19:25, 13 July 2011 (UTC)

They don't get blocked because they are wikilawyering, which means (mis)using the policies rather than breaking them. North8000 (talk) 20:03, 13 July 2011 (UTC)

Maybe they don't get blocked directly for wikilawyering, but an editor who people generally agree is wikilawyering is doing so because they are already on thin ice and they realize it. This policy is not, and should not attempt to be, a behavior policy. Nuujinn is right that "POV Warriors" always end up reveling themselves and that we shouldn't write a specific policy to try to prevent disruption from starting at all.
— V = IR (Talk • Contribs) 20:27, 13 July 2011 (UTC)

The points of the discussion here are: 1. That the wording of the policy can make the situation worse or better. 2. That normal editing practices follow wp;ver but don't follow the other mantras derived from it 3. That, aided by policy wording, those normal practices break down whenever there is a contentious situation. North8000 (talk) 20:46, 13 July 2011 (UTC)

To address these points: 1. You are correct that the wording of a policy can make a situation better or worse... but there is a sub-point: That changing the wording of the policy in an attempt to make one situation better, can result in making another, more serious situation worse (my major concern here) 2. I think normal editing practices follow WP:ver and the mantras derived from it. Non-normal editing occasionally varies from both policy and manra. 3. When there is a controversy, normal editing practices break down regardless of the policy. Blueboar (talk) 23:36, 13 July 2011 (UTC)

A real-life example of conflict between V and T

Here is a real-life example of a conflict between "verifiable" as in it can be verified it exists in a WP:RS, that conflict with "the Truth". The founder of the Subway chain of restaurants and where he came up with the idea- multiple newspaper articles written over 10+ years stuck to the story that he worked at Mike's Neebas (another sub sandwich chain) in Schenectady, NY one summer when his family vacationed there. Well, one day I noticed that he would have been 9 yrs old when that supposedly he was working there, so I did lots of googling and found that yes, his family did vacation there, and the IDEA of Subway did come from him having eaten at Mike's many times during the summer's in Schenectady, but that he never worked there. So one source, contradicting multiple other sources that had an obvious error and truthiness-problem. Current strict application of WP:RS would assume that one would use the multiple sources as "verifiable" over one other source that conflicts with it. How do we write policy so that it would not come to a bad conclussion for this particular conflict?

WP:IAR? Your example is close to original research, but then again, we all do that sometimes. In such a situation, it is best to a) discuss this on the talk page to get other people's opinion, and b) present it in the article like "while most sources state that he did X (ref 1, 2, 3), he was only 9 years old at the time and he got the idea from Y (ref 4)". If you don't have ref 4, it wouldn't be acceptable, buy with it, it seems alright to me: you give more weight on a minority of sources based on common sense, but you do include the version of the majority of sources. I've had a similar problem with the name of a region, where a source stated that it first appeared in 1844, but many older books used the term (verified through Google books and so on). Because I had no secondary source to support that, only primary sources, but beacuse on the other hand the one secondary source we had was clearly wrong, we eventually excluded the claim altogether. "Verifiability, not truth" is a general rule, not a suicide pact, and editor's came come to different conclusions, but if the choice is between a verifiable claim and an unverifiable claim (e.g. a family member contradicts all reliable sources about his relative, based on personal knowledge), the verifiable claim will trump the unverifiable "truth" (which may be true, or not, but we have no way of knowing this). In such a case, we will not even include the unpublished claim from personal knowledge: it is not acceptable and fails WP:V as it is written now. Fram (talk) 06:46, 14 July 2011 (UTC)

What about this hypothetical–I am convicted of whatever crime that gets national attention (treason or something similarly rare that gets picked up nationally) and the Associated Press spells my last name wrong, or puts my wrong birthday, or gets some other fact wrong about me that would be put into a Wikipedia article on me. While on death row with nothing better to do I somehow have access to internet and complain here about whatever. Since the AP is used by pretty much every news outlet that is a secondary or tertiary source how do we rectify the problem of verifiable sources (all ultimately using the same source as theirs) being simply wrong about a fact. In this scenario we assume any outlets getting the information correct are either unpublished or hard to verify (tv and radio are hard to cite in Wikipedia and often dont go into detail like newspapers on things like birthday and place of origin).Camelbinky (talk) 18:28, 14 July 2011 (UTC)

I wonder if emailing your family and asking them to send a copy of your birth certificate through to something like OTRS would do? A birth certificate should count as a RS, I'd think :o) Of course, that wouldn't work at all if it was something other than your name / date of birth which was wrong. Pesky (talk …stalk!) 04:44, 17 July 2011 (UTC)

A suggestion

Since I've started watching this talk page, I've been having a daily reaction of tl;dr, but I appreciate that "not truth" remains an unresolved issue in many editors' minds. I've said before that I feel strongly that I want to keep "verifiability, not truth". I'd like to suggest something concrete that (maybe?) might help resolve this issue.

Currently, the page begins:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true.

I suggest adding another sentence after it:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Wikipedia strives to present what those reliable sources say in as accurate a way as possible.

My hope is that the second sentence puts to rest any implication that we want to knowingly say something that is not true. Thoughts? --Tryptofish (talk) 19:49, 14 July 2011 (UTC)

I've never liked that the wording is "whether readers can check that material" because it isnt our job (and policy is quite clear on this point) to make sure that ALL readers CAN check, only that SOMEONE can check. Perhaps that wording is why we have the perennial discussions regarding sources in English, pay-sites, real-life sources as opposed to being online, and obscure locations of sources that can only be accessed by someone in that area (a museum for instance). Perhaps we can rectify that issue at the same time. As for your change, I dont think it gets the point across enough about the difference between verifiability and truth. Too much opposition exists for removing that phrase unfortunately, but how about–

The threshold for inclusion in Wikipedia is verifiability, not truth–whether someone is able to check that material in Wikipedia has already been published by a reliable source, and not the opinions of individual editors. Wikipedia strives to present the most accurate knowledge of a subject as published in reliable sources.Camelbinky (talk) 20:37, 14 July 2011 (UTC)

I like tryptofish's added sentence (it's more... pithy ), and I like your change to the first (that it's someone, not absolutely anyone). So: "The threshold for inclusion in Wikipedia is verifiability, not truth–whether someone is able to check that material in Wikipedia has already been published by a reliable source, and not the opinions of individual editors. Wikipedia strives to present what those reliable sources say in as accurate a way as possible."
— V = IR (Talk • Contribs) 21:04, 14 July 2011 (UTC)

That's really good stuff! Both Tryptofish's sentence and the update to it by Ohms & Camelbinky :o) Nice and clear, dead easy to understand by anyone at all, and sums it up beautifully. Pesky (talk …stalk!) 09:28, 15 July 2011 (UTC)

Thanks all of you for those comments! I'd like to tweak this a bit more, basically taking what you have done with the first sentence and trying to keep the syntax consistent between the middle and last parts of the sentence, and also trying to avoid the vagueness of the word "someone", which makes me want to ask "who?". So:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether one is able to check that material in Wikipedia has already been published by a reliable source, not whether individual editors think it is true. Wikipedia strives to present what those reliable sources say in as accurate a way as possible.

--Tryptofish (talk) 20:11, 15 July 2011 (UTC)

The construct "whether one is able" is rather... stilted.
— V = IR (Talk • Contribs) 23:11, 16 July 2011 (UTC)

Thanks for replying! Yeah, I see what you mean. That came from the stuff about getting away from "readers can check". How about:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether anyone is able to check that material in Wikipedia has already been published by a reliable source, not whether individual editors think it is true. Wikipedia strives to present what those reliable sources say in as accurate a way as possible.

What I did was to go back to the wording that you proposed, but change "someone" (which made me wonder "who?") to "anyone". I'd even go for "anyone can check" instead of "anyone is able to check". --Tryptofish (talk) 19:41, 17 July 2011 (UTC)

We need a good "middle ground" effort such as this. But I'm afraid this will be lost / unnoticed on the talk page being farther back. North8000 (talk) 20:10, 17 July 2011 (UTC)

Thanks, me too. Let's try to get some wording that the few of us paying attention can feel good about, here in this relatively quiet space. Then, maybe tomorrow, I'll move it to the bottom as a proposal (as opposed to as a suggestion). --Tryptofish (talk) 20:15, 17 July 2011 (UTC)

I was going to suggest that we move that to a subpage for quiet, as I think North8000 is correct. But this sounds like a plan. --Nuujinn (talk) 20:18, 17 July 2011 (UTC)

Would it make sense to change "published by a reliable source" to "published by one or more reliable sources"? --Nuujinn (talk) 20:20, 17 July 2011 (UTC)

I don't feel strongly either way, but it's part of a blue link, which maybe (?) argues for brevity. And if it's published by one or more, it's de facto published by one. It would also be OK with me to go with "published by reliable sources". --Tryptofish (talk) 20:37, 17 July 2011 (UTC)

Speaking only for myself, something that moves this paragraph to overall neutral ground regarding accuracy would settle it for me. The issue is that as worded, it is interpreted as say that seeking accuracy (in cases where objective accuracy exists) is not a legitimate activity for editors, or a legitimate consideration. A non-structural statement that merely shifts the psychology of this paragraph to neutral ground on that aspect would do it for me as a compromise. Something as simple as "Verifiability helps achieve accuracy" would do it for me. Folks on one side would say it's not enough, folks on the other side would say it's heresy compared with mantras erroneously mis-derived from wp:ver. Such is the fate of compromises and middle grounds.  :-) Sincerely, North8000 (talk) 20:48, 17 July 2011 (UTC)

How about:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether one is able to check that material in Wikipedia has already been published by one or more reliable sources, not whether individual editors think it is true. Verifiability helps ensure accuracy, and Wikipedia strives to present what reliable sources say in as accurate a way as possible. --Nuujinn (talk) 23:58, 17 July 2011 (UTC)

Not sure what I think about that one. It's sort of a compromise to the compromise. But I think I could go with it. North8000 (talk) 11:24, 18 July 2011 (UTC)

additional discussion

Oh, Oh... or maybe we could say:

• The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. The fact that an editor or editors are 100% positive that something is true is irrelevent; if it is not verifiable (meaning not published by a reliable source) you can't include it. Seriously... we mean this. We honestly don't care how true you think it is... it has to be verifiable to be included. and by that we mean if you can't point to a reliable source for it, don't add it! Really.

Wait... I think I saw ... nope, sorry, my mistake, the horse is still dead... maybe if we tried beating it with birch rods instead of hickory sticks? sorry for the sarcasm folks... but this is getting ridiculous... we are rehashing stuff we have been over time and again. No one is budging and it really is time to move onBlueboar (talk) 02:12, 17 July 2011 (UTC)

I still like "The threshold for inclusion in Wikipedia is Verifiable truth, not unverifiable truth" followed by the "The fact that an editor or editors are 100% positive that something is true is irrelevent; if it is not verifiable (meaning not published by a reliable source) you can't include it. Seriously... we mean this. We honestly don't care how true you think it is... it has to be verifiable to be included. and by that we mean if you can't point to a reliable source for it, don't add it! Really" thing - all it's doing is swapping "Verifiability, not truth" out for "Verifiable truth, not unverifiable truth".

Would anyone who just swapped that in actually get shot down in flames for doing it? Is there anything actually wrong with that version instead? Is is actually worse? Would you kill me if I bunged it in? Could we discuss that particular version as a possible improvement, rather than go round in circles again? Love you all, guys, really I do (granny speaking here!) ..... but let's try to take a step forwards, rather than backing away from the carcass again (it might only be under anaesthetic, not dead) :o) Pesky (talk …stalk!) 04:52, 17 July 2011 (UTC)

Please don't, and please let us finally drop discussion of the first line for some reasonable period of time. This is simply exhausting--do you not see how very long this has been going on, here and at the pump? If you want to proceed, I ask that you reread the entire discussion in the archives in both venues. If you're a fast reader, I expect to see you back in a week or so.... (; --Nuujinn (talk) 13:33, 17 July 2011 (UTC)

The problem I have is not just the "not truth", but the combination of this with the "has been published in a reliable source". This means that if I know that something is true and it is verifiable in the usual scientific meaning, I could still not write about it on Wikipedia, if there is no single source from which you can directly verify it. This is at odds with how technical scientific wiki-articles are edited. Count Iblis (talk) 14:52, 17 July 2011 (UTC)

I'm not quite sure how. Can you point to an example, or sketch one out?
— V = IR (Talk • Contribs) 15:22, 17 July 2011 (UTC)

It is actually closely related to what I wrote here. You can't always verify some given explanation from a single source. Suppose that the given derivation of the quantum version of Poincaré's theorem were challenged, that argument would not be settled by merely pointing to out some references. One woud have to engage in an argument about why the Wikipedia text is an acceptable or not acceptable way to present the mathematical argument that is given in the cited sources. Count Iblis (talk) 17:16, 17 July 2011 (UTC)

I disagree. If those disagreements were in primary sources, we shouldn't be using them in the first place, unless they were backed up by reliable secondary sources. If the arguments were presented in a secondary source, we would document what the secondary source says. We're writing an encyclopedia, not evaluating mathematicians, that would violate OR. That's what academic journals are for. --Nuujinn (talk) 17:59, 17 July 2011 (UTC)

The "disagreement" here would be about verifying a Wiki-text. So, if you have written up a paragraph about some mathemtical subject and you have paraphrased also the mathematical arguments, then verifying that your text correctly represents the references may involve non-trivial math. If there are questions raised about that, then that may not always be settled by pointing to one ref, it can be a matter of detailed arguments and perhaps some refs to textbooks. Count Iblis (talk) 18:10, 17 July 2011 (UTC)

And what you are talking about sounds very much like a WP:SYNT violation... a form of Original research. Blueboar (talk) 01:06, 18 July 2011 (UTC)

Yes, it is not our job to verify the work of academicians. Reporting what a source such as wiki entry written by an acknowledged expert is fine, but if you have to resort to your own calculations or combine other information from other sources to check your math, you've most likely crossed the line into OR. If you are competent to do that kind of work, doing so is more appropriate elsewhere. --Nuujinn (talk) 01:14, 18 July 2011 (UTC)

This does not happen when you write such an article, it's only an issue when someone who doesn't have the necessary background to understand the source, wants to verify the content. This is because paraphrasing the math in a way that is routine from the POV of someone who does understand the source, may not at all be routine from the POV of someone who doesn't. Then because such an editor should still be able to satisfy him/herself that nothing improper is going on w.r.t. OR etc. etc., that should be settled by discussions on the talk page.

But in such discussions, it would then not be ok. for that editor to demand direct citations. Simply put, editors who lack the knowledge to understand the cited refs themselves should not complain about Synth, when others try to fill in the gaps in the knowledge of the editor to let that editor understand it. Count Iblis (talk) 15:13, 18 July 2011 (UTC)

As an example of what Count Iblis is saying, consider star positions. In any one astronomy book or article, it is normal to give the positions as of a consistent date. A star catalog might give the positions as of January 1, 2000. The Astronomical Almanac gives the position as of the middle of the year of the almanac. Converting the position from one epoch to another is too hard for someone with a casual interest in astronomy to do (which might describe many of our readers). But it is too easy to be worthy of publication in an astronomy journal or book (the method might be published, but not the specific results). So there will never be a reliable source that could be cited to prove that a specific conversion in a Wikipedia article was done correctly. The best that could be done would be to cite reliable computer software that would do the conversion. Jc3s5h (talk) 15:31, 18 July 2011 (UTC)

How about this? As a compromise, develop a sentence that follows that addresses the main expressed concerns, and then drop the discussion of the first sentence for 1/2 year? North8000 (talk) 15:26, 17 July 2011 (UTC)

Float some suggestions... perhaps you will come up with something we all like. However, (Speaking purely for myself and not for others) I really like our current language and I have yet to see a suggestion that I think is better. That does not mean better language is impossible, but (being completely honest) it will be a hard sell. Blueboar (talk) 01:06, 18 July 2011 (UTC)

Agreed. I get the sense that I'm (slightly) more willing to consider additions then you are, but... Anything short of substantially altering the first sentence, I'm certainly willing to listen, if not outright support. The "someone"/"anyone" kind of question above is one thing worth considering, for example, but it does need to be pitched well and considered.
— V = IR (Talk • Contribs) 02:08, 18 July 2011 (UTC)

I know better wording will be a hard sell, but I'm still willing to try and work on it. I'm sure everyone here would agree that better wording (by definition!) would be better - and improvement is always worth working towards, no matter how bloody fed-up we are with discussing it! I've read, and read, and read .... (I'm an exceptionally fast reader!) There will be a way of getting a breakthrough moment - one of those "AHA!" things, provided that we don't actually quit. Of course, nothing stopping those who want to quit from quitting, and those who'd really like to improve it from carrying on working towards improvement - we're all volunteers, after all. We all hit times when we just don't want to consider anything at all - we get emotionally bogged down in stuff; of course we do, we're human! But there is a better way, out there, somewhere. It's just up to [probably some of] us to seek it out. Pesky (talk …stalk!) 08:17, 18 July 2011 (UTC)

A way to move forward

OK, it sounds like we actually may have some consensus here... not on language, but on approach. Here is how I think we currently stand:

• There is strong opposition to changing the first part of the first sentence ("The threshold for inclusion in Wikipedia is verifiability, not truth—"

Because of this opposition, we should drop the idea of changing it... at least for now.

• There is serious hesitation about changing the second part of the first sentence ("—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true.")

This second part of the first sentence explains what we mean in the first part. A small majority of us really like this language, and are very reluctant to change it. However, if someone can convince us that some other language is actually better, we are willing consider it. It's going to be a hard sell.

• There is guarded acceptance of the idea of adding additional explanatory language... adding a (yet to be written) second or third sentence to further explain the first sentence.

From my perspective, this last is the best approach. It still won't be easy... but since we are not "removing" what already exists, merely expanding on it, there will be less resistance (and more chance of reaching a consensus on the language).

If people agree with my assessment... perhaps we can move to the next step... proposing and discussing potential additions. Blueboar (talk) 15:12, 18 July 2011 (UTC)

Are you suggesting a specific compromise, like drop the idea of changing it, and agree that we'll add a sentence that addresses the expressed "not truth" concerns? Or is it to saying to drop it in exchange for the possibility of a hope of getting the latter? North8000 (talk) 16:02, 18 July 2011 (UTC)

Neither. I included the caveat of "at least for now" merely as a recognition that opinions could (perhaps, maybe, who knows, unlikely but stranger things have happened) change at some vague point in the future. My point is this... right now, the reality is that there is no fucking way we are going to achieve a consensus to change the first part of the disputed sentence... and very little chance that we will achieve a consensus to change the second part. However, it looks like there is a chance of achieving a consensus on adding some sort of further explanatory language ... so ... I am suggesting that rather than continue to but heads over something that we all know we don't have a consensus on (and probably won't for the foreseeable future), we should focus on doing something that we actually are likely to achieve a consensus on.... finding some sort of acceptable additional language that we can all agree with. Blueboar (talk) 17:24, 18 July 2011 (UTC)

I actually really liked your earlier suggestions of ""The fact that an editor or editors are 100% positive that something is true is irrelevent; if it is not verifiable (meaning not published by a reliable source) you can't include it. Seriously... we mean this. We honestly don't care how true you think it is... it has to be verifiable to be included. and by that we mean if you can't point to a reliable source for it, don't add it! Really!" It's absolutely dead clear. Pesky (talk …stalk!) 17:25, 18 July 2011 (UTC)

In case you didn't notice... I was being sarcastic when I said that. We don't need to repeat the same thing six of seven times to get the point across. Blueboar (talk) 17:40, 18 July 2011 (UTC)

Blueboar, I'm cool with that. Basically simply focus on the possible addition and see where that goes. North8000 (talk) 19:53, 18 July 2011 (UTC)

There are quite of few sentences which have been proposed as possible additions: see the above discussion Wikipedia_talk:Verifiability#A_suggestion. Most of them look pretty good to me. Mlm42 (talk) 20:01, 18 July 2011 (UTC)

"not truth" is a Figure of speech

The Wikipedia article Figure of speech has theory that helps to understand the problem in the first sentence of the Project Page with not truth.

A figure of speech is the use of a word or words diverging from its usual meaning.
. . .
Figures of speech often provide emphasis, freshness of expression, or clarity. However, clarity may also suffer from their use, as any figure of speech introduces an ambiguity between literal and figurative interpretation.

Thus, those defending the use of this figure of speech are also telling the world that we the writer's of this encylopedia lack sufficient skill in writing to write our policies without ambiguity.  Unscintillating (talk) 01:02, 14 June 2011 (UTC)

No... it isn't a Figure of speech... we actually mean what we say. Like it or not, truth is not the threshold for inclusion in Wikipedia... "but, it's true" has never been an acceptable reason to include material. We require verifiability. Blueboar (talk) 15:40, 14 June 2011 (UTC)

By claiming that not truth is not a figure of speech, Blueboar, you self-identify as a "true believer in WP:V".  Unscintillating (talk) 03:05, 15 June 2011 (UTC)

4500 totally unreferenced BLPs say you're wrong. SBHarris 01:44, 15 June 2011 (UTC)

We have to be more careful in our phrasing and our explanation. We are doing something other than saying "truth is not the threshold for inclusion", we're saying 'not truth', at the end of an unfamiliar phrase (threshold for inclusion). To clearly explain the meaning of a critical phrase is a good idea. Ocaasi ^{t | c} 20:41, 14 June 2011 (UTC)

It's actually a phrase that is put into a sentence about another topic (verifiability requirement), intended to strengthen the verifiability requirement, but which, as written, is prone to getting mis-quoted and mis-applied. North8000 (talk) 21:01, 14 June 2011 (UTC)

Is there a debate here? I am with Blueboar here. Slrubenstein | Talk 21:25, 14 June 2011 (UTC)

If you are with Blueboar then you believe that we mean what we say literally here?  Why is Jimbo quoted on the WP:V page, "Zero information is preferred to misleading or false information"..."I can NOT emphasize this enough."  Why does WP:Editing policy say, "...on Wikipedia a lack of information is better than misleading or false information—Wikipedia's reputation as a trusted encyclopedia depends on the information in articles being verifiable and reliable."?  What does the five pillars mean when it says about WP:V, "All articles must strive for verifiable accuracy"?  Are the WP:5 fundamental principles wrong?  Unscintillating (talk) 03:05, 15 June 2011 (UTC)

I don't see a contradiction between Jimbo's statement and Blueboar's.. actually, I'm somewhat surprised this discussion hasn't degenerated into a philosophical discussion about the words "truth", "knowledge", and "information".. these words mean different things to different people. Mlm42 (talk) 03:37, 15 June 2011 (UTC)

So you are claiming that Jimbo thinks that verifiable material that may be not true is acceptable in Wikipedia, and that Blueboar's acceptance of material that may be not true is material that is neither misleading nor false because it is verifiable?  Unscintillating (talk) 03:57, 15 June 2011 (UTC)

No, Jimbo was saying that we shouldn't be including something which is known to be untrue. You're confusing the concepts a bit; in particular the phrase "verifiable material" is being misused. Example: "X says the Earth is flat" can be a verifiable (and true) statement, even though "the Earth is flat" is not a true statement. Reliable sources can be mistaken about things.. reliable sources disagree.. we don't actually have to know the "truth" to write an encyclopedia. Mlm42 (talk) 04:19, 15 June 2011 (UTC)

Please soften the personal characterizations, it might help to add "IMO", IMO.  Unscintillating (talk) 06:35, 15 June 2011 (UTC)

I think that True believers in WP:V don't think that editors should spend time to consider if something verifiable may be not true, that their time is protected by WP:V; in other words, that Jimbo can't claim something is false without POV truth pushing.  Unscintillating (talk) 06:35, 15 June 2011 (UTC)

Also, that is a fallacy, that we can blindly write an encyclopedia without considering the accuracy of our sources.  A specific case, WP:ELNO requires us to know if the source is misleading.  Unscintillating (talk) 06:35, 15 June 2011 (UTC)

It's not a figure of speech. The sentence means this (probably called a "compound predicate" or some such to the grammar mavens):

• The threshold for inclusion is verifiability.
• The threshold for inclusion is not truth.

This is not complicated. The construction works exactly like "The book is red, not blue." That sentence also divides in the same manner:

• The book is red.
• The book is not blue.

Nobody would look at this sentence, which uses exactly the same grammar, and conclude that the speaker is opposed to books with blue covers. It's equally silly to look at "verifiability, not truth" and conclude that Wikipedia is anti-truth. WhatamIdoing (talk) 04:50, 15 June 2011 (UTC)

The problem is that what the sentence intends to mean is that the "theshold" to inclusion is desired to be(truth+verifiability) rather than (truth-only). It's easily misunderstood to mean (lie+verifiability) rather than (truth-only). Or (who-cares-whether-true-or-not + verifiability) rather than (truth-only). Which rankles.

In addition, there is the serious problem that policies are supposed to be descriptive rather that prescriptive, and a description of how WP is actually written, is that the inclusion threshold is indeed truth-only. HENCE the many, many articles (including those 4500 unreferenced BLPs) that have no citations at all. Most of Wikipedia (if you look at the info bit-by-bit) is actually still-unreferenced. Most of it continues to survive because a lot of people have looked at it and agreed with it as being "true," and left it alone even though still unverified (or at least uncited). That is how WP really actually operates. However (again) if by "verifiable" you mean "verifiable in theory" even though still uncited, that needs to be said up front and early. SBHarris 05:51, 15 June 2011 (UTC)

On "actually operates" it's amazing how few people notice that gorilla in the living room. North8000 (talk) 10:54, 15 June 2011 (UTC)

Sbharris, you seem to be confusing verifiABLE with verifiED. Unreferenced articles are frequently 100% verifiABLE. Material is verifiABLE if the source exists—that is, if an appropriate reliable source has been WP:Published somewhere in the world, even if zero Wikipedians have ever written down what the source is. Citations do not make material verifiABLE (which is what's required as the minimum standard for including information). Naming your sources makes the material verifiED. WhatamIdoing (talk) 16:32, 15 June 2011 (UTC)

Um, that distinction belongs in line one, rather than waiting for lines three and four. "readers can check (in theory) that material in Wikipedia has already been published by a reliable source.." So "you could look it up" as we say in arguments, even if we don't know where. And in addition, the "not truth" needs to be qualified to mean "not ONLY truth."

The threshold for inclusion in Wikipedia is verifiability plus truth, not simply truth. This means that readers could (in theory) check that material in Wikipedia has already been published by a reliable source, not simply whether editors think it is true.

SBHarris 21:31, 15 June 2011 (UTC)

So where do I sign up to delete all that climate change stuff? You're actually proposing that that information must be both verifiable and True™ before we can include it. Well, I say climate change is not true, and that that there's no debate over abortion rights, that the riots in the Middle East aren't happening, and that Taiwan isn't separate from China, so Wikipedia shouldn't include any of that, according to your proposal.

This obviously isn't going to work. We cannot allow editors to decide for themselves whether something is True and only include what they happen to believe. WhatamIdoing (talk) 21:22, 17 June 2011 (UTC)

I think that in the part of Wikipedia that works (non-contentious topics) that when there is a question, it's very common for the editors to decide which choice IS true/accurate, put in in and then source it. In the part of Wikipedia that doesn't work (contentious articles) this process breaks down; there either is no consensus, or somebody uses Wikilawyering to knock out the result. North8000 (talk) 21:32, 17 June 2011 (UTC)

So it is "not complicated"?  Is not the phrase "not truth" used to disparage people who come forward to represent the "truth"?  Is not the word "truth" there turned into mockery and sarcasm?  Does the word "truth" then "diverge from its usual meaning"?  Why did SV say earlier that not truth is "iconic"?  Does this not come from the "emphasis" of a figure of speech?  So why claim that it is not a figure of speech?  Please support removing this figure of speech from this technical writing.  Thanks, Unscintillating (talk) 07:00, 15 June 2011 (UTC)

It's equally silly to look at "verifiability, not truth" and conclude that Wikipedia is anti-truth. -- It's silly, because it's already clear from a cursory glance at almost all articles about any hot button topic. --213.196.218.59 (talk) 14:51, 15 June 2011 (UTC)

• If this was actually the problem, then we could say, "The threshold for inclusion is verifiability rather than truth." But to focus on the exact syntax, as this proposal does, is a misunderstanding of the problem.—S Marshall T/C 11:30, 15 June 2011 (UTC)

On the articles that work, the editors seek accuracy. On the articles where the current policies are an abysmal failure (basically all of the contentious articles) "not truth" is used to disparage the concept of accuracy when such will serve the POV warrior's purpose. North8000 (talk) 16:47, 15 June 2011 (UTC)

North, much as I support your intentions, changing this phrase won't end edit-warring, where people have different versions of what is accurate, and relying on sources is the only way to go forward. That said, we should emphasize in this policy that: 1) truth on Wikipedia is ascertained through a balance of the best sources; and 2) threshold means there are secondary considerations. Ocaasi ^{t | c} 17:09, 15 June 2011 (UTC)

As a result of the context, my comment overstated the role of the first sentence in the problem. North8000 (talk) 17:26, 15 June 2011 (UTC)

• The addition of "truth" a bad idea. We can agree (in principle) on what is verifiable; we can not expect (at least according to most theories of knowledge) to agree on what is truth. To nclude the word "truth" is a invitation to endless dissension and conflict for politics, religion, and countless other things also. The principle of NPOV, equally important to V, is that we leave the determination of truth up to the reader. TThe second phrase may be intended to explain the first, as saying that we consider ruth as being able to potentially trace all information to its source. But that is what Verifiability means all by itself. This attempt to modify the present wording weakens it. Some things we have gotten right. DGG ( talk ) 00:48, 18 June 2011 (UTC)

• DGG, the point here is that it's a sin against the basic purpose of an encyclopaedia to publish known error except to refute it. We tend to get this right in practice, for example in Climate change controversy—there are "reliable sources" that say anthropogenic climate change is a myth, but Wikipedia rightly decides to follow the scientific consensus rather than the media. Despite WP:NPOV, which can be summarised as "Wikipedia describes disputes but does not pick sides in them", when the chips are really down (such as in Climate change controversy or Young earth creationism) we choose to tell the truth even though there are reliable sources that disagree.

  This is usually described in terms of selecting the most reliable source, but the logic there doesn't work: how do we select the most reliable source except by deciding which one is the most credible, the most believable, the most likely to be right? And how do we decide that without deciding what's true?

  The reality is that there is such a thing as objective truth, and in cases like the ones I've mentioned, it matters.—S Marshall T/C 18:27, 18 June 2011 (UTC)

• But, and there is admittedly some irony in this, if we added "truth" the result would be editors pushing the non-scientific version of climate change using that language to claim that they were writing the "truth". --Tryptofish (talk) 19:01, 18 June 2011 (UTC)

• And we round the bend again. Anything that can be considered "objective truth" (and to be clear, I follow Pragmaticism on this) should be easy to source. How we rank sources to come to consensus varies from topic to topic, and are arrived at through discussion. For climate change, we use scientific sources. For other kinds of information, different kinds of sources are more appropriate. --Nuujinn (talk) 19:09, 18 June 2011 (UTC)

• Nuujin, you're right to say that "objective truth" is easy to source. Unfortunately, blatant lies are also easy to source, and sometimes, subtle lies too. ("A lot of the books and articles the contributors cite turn out to be no more reliable than Wikipedia itself."—Geoffrey Nunberg.) The basic problem with the phrase "verifiability, not truth" is that taken at face value, it allows almost any sourced claim to be made in an article and given weight. We do tend to get this right in practice, as in the examples I've cited, but we get it right by selectively disregarding WP:V and WP:NPOV.

  "Verifiability, not truth" is a soundbite and a good put-down for certain POV-pushers. I understand its attraction. Unfortunately, it's not how we really do things. In the really controversial material we pick and choose which source to follow, and we can only do that by deciding what can be trusted—and hence, what's true.—S Marshall T/C 19:41, 18 June 2011 (UTC)

Understood. But the problem is that we still need to be able to respond effectively to those POV pushers. Somewhere above, I suggested that we retain what you call the soundbite (and I think there is consensus that we should retain it), but that we also follow it with an explanatory sentence to address the issue to which you and others refer. It seems that other editors want a break from discussing further changes for a while, but I think we should revisit this possibility when we are ready. --Tryptofish (talk) 19:48, 18 June 2011 (UTC)

• I don't think there's any real consensus about the phrase "verifiability, not truth" at all. I wonder if the answer isn't to have a section about dealing with conflicting sources—which is conspicuously lacking from policy at the moment. WP:NPOV skirts closest to the issue at the moment, but I would think deciding between sources is more of a matter of verifiability than neutrality.—S Marshall T/C 19:57, 18 June 2011 (UTC)

What about a WP:Science policy (it currently redirects to Wikiproject science) modeled after WP:BLP? The BLP policy makes it very clear that accuracy (a wiki-politically correct word for "truth" :) ) is very important. So, one could also say that any Wiki-article in which a claim about a scientific fact is made must be an accurate reflection of what peer reviewed sources actually say on the topic.

This would then cover pretty much all the Wiki-articles where "truth" is more important than this policy page suggests. So by enforcing that policy one would intervene in the correct way while leaving the other sectors of Wikipedia, where "truth" means something else, alone. Count Iblis (talk) 20:17, 18 June 2011 (UTC)

• I don't know about that, Count Iblis. It seems to me that we have two established routes to truth: science and law. But to me, that seems tangential to the difficulty I'm trying to solve. For me, "verifiability, not truth" is fine except in the situation where two opposing or mutually contradictory viewpoints appear in different reliable sources. (I'll mention my examples of young earth creationism and climate change controversy again here, because they clarify the sort of situation I'm getting at, and hopefully reduce the risk of misunderstanding.) Where there's a tension between reliable sources we might describe the tension, or we might come down on one side or the other on the basis of which source seems more trustworthy or more closely in alignment with the evidence (which means we decide which is true). In other words, for me, the problem is more about mutually contradictory sources than it is about science.—S Marshall T/C 21:01, 18 June 2011 (UTC)

And there are other things (the birther controversy comes to mind), where there are differing versions of "truth", with blatantly different degrees of factual support, and which are neither science nor law. --Tryptofish (talk) 21:11, 18 June 2011 (UTC)

That's a matter of the historicity of a single event, which is often very difficult to establish. Legal and scientific tests can be involved (think of forensics) but it's not simply a matter of a combination of the two. And for non-repeating events in the past, archeology doesn't help, either. History per se has its own standards, and those are what are used.

Example: consider the John F. Kennedy assassination, which is history informed by science and law. The consensus of editors was to use the findings of the official government panels (of which there have been three) except IN ONE CASE where we didn't believe them (!). How did that happen? Well, they all three concluded that JFK was shot by Oswald alone. The last HSCA panel, however, concluded that there was recorded acustic evidence of an additional 4th shot fired too close to the others to have been Oswald, and thus a second gunman, and thus a conspiracy. However, they could find NO other good evidence for a conspiracy, though they looked harder than anybody has before or since. That last sound (recorded on a dictabelt) was the ONLY evidence for a fourth (missed) shot. After that commission disbanded a lot of good evidence turned up that the recording was NOT of the assassination, and thus the fourth shot, the second gunman, and the only good conspiracy evidence all vanish. Except that the government panel was not about to reconvene to admit that it had erred. So WP's editors decided to make the official WP version that Oswald had acted alone, but mention the screwed-up third panel and the later people who disagreed with this ONE finding of it (but not its many others). However, none of this consideration was given to people who disagreed with the earlier panels (Warren and Clark panels).

Now, all this may seem odd or hypocritical or not playing by any given set of rules. But that is what the majority of editors on this article think happened to JFK, i.e., what was true. It's probably also the conclusion of the majority of historians of the event, save those with an alternative theory and a book to sell. And alas, due to Oliver Stone, the general public doesn't agree with it, either! So alternate theories have been subbed (don't say forked!) off to their own subarticle John F. Kennedy assassination conspiracy theories.

Now, I happen to agree with the conclusion that Oswald acted alone, so I think WP got it right. However, somebody who does not agree with me is going to be very unhappy. All of this raises difficult issues about truth on WP, and how standards of truth-finding are arrived-at. Basically, there aren't any that are universal, especially for history. NOPV coverage of an event like the JFK assassination is basically impossible, do don't even suggest it. One steers by authority and (in a few cases) by later events that the authorities missed because they never addressed them, but many others have picked up since. But that narrative thread is based on editor's judgement. You can always find as many cites as you like for other views. SBHarris 22:42, 18 June 2011 (UTC)

The Kennedy asassination is probably a bad example. A rare borderline case where a fringe theory has more than the normal amount of traction/believers, and slightly-credible material written about it. North8000 (talk) 15:12, 19 June 2011 (UTC) And the ones where there clearly is no objective reality (e.g. is Obama a good president?) are also irrelevant. Here are two involving objective reality that are a little further from that line:

• Global warming: Did we cause the rise that has occurred? Strong but not conclusive evidence says we did, many, and many rs's say we didn't.

• Something other than an airliner crashed into the Pentagon on 9/11. Clearly a false assertion, but wp:"rs's" can be found covering claims on both sides. North8000 (talk) 18:12, 19 June 2011 (UTC)

• You don't even need to get into conspiracy theory. For millions of people, the statement: "Jesus is the Son of God" is held to be absolutely 100% true. Tons of reliable theological sources support the statement. So... can we add that blunt, unattributed, unhedged statement to an article? Of course not. Despite the fact that millions of people consider the statement to be absolutely true, there is no way to verify it (or to put it another way... the divinity of Jesus Christ is a matter of belief rather than verifiability... it can only be accepted as truth or rejected as untruth). What is verifiable however is the statement: "Millions of Christians believe that Jesus is the Son of God". Blueboar (talk) 20:34, 19 June 2011 (UTC)

• Contrary to what you say, Blueboar, in a Wikipedian sense, it's "verifiable" that Jesus died for our sins, and then rose from the dead. This is because reliable sources have published that, and Wikipedia explicitly draws a distinction between verifiability and truth--which means that "truth" is not part of Wikipedia's definition of "verifiability". In other words, thanks to the phrase "verifiability, not truth", irrespective of whether it's true that Jesus died for our sins and rose from the dead, you can still write it on Wikipedia provided you can cite a source. In practice Wikipedia doesn't work like that because editors will make you attribute your statements to your sources if they don't think what you say is true--so on Wikipedia, you can write "Millions of Christians believe that Jesus died for our sins and rose from the dead"--but this is despite what policy says, and not because of it. But you can write "the Universe was created about 13.5 billion years ago" without attributing it to a source. (Interestingly, on Conservapedia these situations would be reversed, so you could write "Jesus died from our sins and rose from the dead" quite baldly, but you would have to write "Atheists claim the Universe was created about 13.5 billion years ago".)

  I think the way Wikipedia handles it in practice is a good thing, in general, but I also think it's a problem that in this case our policy doesn't describe our behaviour.—S Marshall T/C 20:49, 19 June 2011 (UTC)

A very accurate summary. But the missing qualifier on your "in practice" statement is that, on contentious articles, anyone can and does capitalize on that conflict between policy and reality and wikilawyer to declare the normal Wikipedia practice to be illegal. North8000 (talk) 21:56, 19 June 2011 (UTC)

Which they will do no matter what we say in the policy. If we remove the "not truth" clause, then the POV pushing wikilawyers will argue that unverifiable material may be added on the grounds that it is "true" (ie someone believe it to be true)... on the other hand, if we leave it in then the POV pushing wikilawyers will argue that marginally verifiable material may not be removed on the grounds that it is "not true" (ie someone believes it is not true). However, because we can limit the addition of marginally verifiable "not true" information (through application of WP:UNDUE, WP:NOR and other policies), the second scenario (while annoying) is less damaging to the project over all. Blueboar (talk) 13:37, 20 June 2011 (UTC)

I'm burnt out on talking about the first sentence. I was viewing this as a useful broader discussion. I was seconding / addressing the more general topic on how the editing process that has created much of Wikipedia's content bites the dust when the article is contentious. The first sentence of wp:ver is only one of many contributors to the problem, and the main causes / fixes relate more to what isn't written in policies than what is. Sincerely, North8000 (talk) 13:54, 20 June 2011 (UTC)

• I don't think we're really talking about the first sentence. I think we're talking about how to handle conflicting sources. We don't need to reinvent the wheel. All that's needed is for policy to document what we already do in cases like young earth creationism or climate change controversy. This will probably entail some, er, clarification of the words "not truth" but that's incidental—a by-product of an attempt to address the real issue.—S Marshall T/C 14:01, 20 June 2011 (UTC)

Are we meant to be building an Encyclopedia or not?

In science articles this is not so much a problem as usually the most recent published works are going to be the most accurate. However in the humanities this is not the case. There are thousands of published peer reviewed books and papers out there which regurgitate the same old mistakes (often with political agendas) again and again. Information is not knowledge and the job of an encyclopedia is to cut through the info to furnish readers with knowledge. We live in an age of information and one has to be politically savvy in order to be able to cut through the spin and political agenda to find the knowledge behind the info. There has to be a way to furnish our readers with critical reportage of these views not just regurgitating them as if they were fact when there are clear problems in relation to the earliest sources. Let's take history as an example. It is very possible to build up an argument and another argument upon the speculation of one author on the speculation of another author upon the primary sources. While it is very necessary to report what each author has done, it is also much more important to report upon the original sources exactly as was done in the early Encyclopedia Britannicas which furnished the readers with the bare bone facts before going into the discussions of scholars upon those facts and how theories have developed. This is because in History opposite to science, the primary sources are the evidence which (until time machines are invented) can never be discarded in favor of the most recent theories. History is just one example. It seems a lot of the policies on Wikipedia intended for use in building up science articles are abused when it comes to History articles. This is why there are more edit conflicts in such articles. while science articles grow into very useful and respectable sources of information. Basically people with strong agendas are abusing wiki policies in order to cover up the past and often build a fantasy picture tinted by all sorts of political and ideological and sometimes even religious agendas. If this has all been delt with before, I hope someone can point me to the right place for discussion on this topic. Many thanks.81.103.121.144 (talk) 09:57, 20 June 2011 (UTC)

Well, we are writing an encyclopedia indeed and primarily a general one and if you look at (most) general (print) encyclopedia out there, they simply state (potentially outdated) (traditional) mainstream views rather than what doing what you seem to suggest. While I can see your problem, I don't think there is a simple solution to it. First of all in a addition to (secondary) literature you can provide primary sources as well (for illustration and augmentation) and quote/describe them as you see fit. However what you can not do is providing your own analysis of them but you need to report their analysis in some secondary literature instead. Allowing you to analyze the primaries sources is not fixing the problem of agenda driven falsehoods/misleading representations but worsening it. Because then you move the spinning from external literature directly into WP. Meaning essentially any author might claim anything based on his personal analysis of primary sources (the conspiracy nuts will rejoice). The reason why we require reporting about secondary external sources is, that they do provide an additional filter helping to lessen the problem of agenda driven spinning (the nuttiest agenda driven pieces simply don't get published in reputable sources). Furthermore this way we somewhat restrict the interpretation to qualified professionals (a historian assessing (primary) historical sources) rather than an arbitrary WP editors, who might lack the qualification and context knowledge for an appropriate assessment. --Kmhkmh (talk) 12:35, 20 June 2011 (UTC)

• Would I be right in thinking that this is about Notzrim?—S Marshall T/C 10:58, 20 June 2011 (UTC)

re: "It seems a lot of the policies on Wikipedia intended for use in building up science articles are abused when it comes to History articles." - As a historian, I can sympathize with this complaint. At times it does seem as if our policies are written with needs of science in mind, and not the needs of the humanities. Then I read the complaints of editors who focus on science articles. They often complain that our policies are biased the other way... written with humanities in mind, and are abused in science articles. That makes me realize that the problem isn't with the policy. The problem is with abusive editors. No matter how we write our policies, there are those who will look for loopholes and ways to abuse them. Blueboar (talk) 11:30, 20 June 2011 (UTC)

Yes, but I think that we could tweak them to significantly reduce the problems. North8000 (talk) 13:56, 20 June 2011 (UTC)

I don't disagree... however, we have to keep in mind that by tweaking to resolve one set of problems, we can often (perhaps inadvertently) create another set of problems. That's why policy changes so slowly. We need to be deliberate and think of the big picture. Whenever we are contemplating a tweak, we need to ask ourselves... how will this tweak be abused? Because it will be abused. Blueboar (talk) 14:22, 20 June 2011 (UTC)

I'm very conscious of that; what you just said sounds like what I'm known for preaching in real life. But I guess I would also say that some current policy wording (or lack thereof) reflects having forgotten that, or inability to forsee those consequences. North8000 (talk) 14:41, 20 June 2011 (UTC)

Wikipedia has strengths and weaknesses. Policy has to be more restrictive in those areas in which Wikipedia has systemic weakness, yet less restrictive in other areas. Articles on major overviews are particularly problematic. By this I mean overarching subject areas. Wikipedia has strength in areas that are well-defined. This refers to subject areas in which there is little doubt as to what the scope of an article on that subject area would be. Wikipedia does best when a topic for an article can be easily named and when there can be little wiggle-room as to what that article title refers to—because reliable sources provide definition for that which is the title of the article. Bus stop (talk) 15:01, 20 June 2011 (UTC)

I agree but this is also simply due to the fact that WP compiles the knowledge of academia and the public. So any issues or topics not really settled or canonized in academia and/or the public (though the former matters more than the latter) cannot be expected to be or appear settled in WP. WP follows the (standardized) knowledge creation rather than running ahead. As long as academia still repeats/publishes outdated or misleading notions, they will be found in WP as well. They will only vanish in WP once they've essentially vanished from academia.--Kmhkmh (talk) 19:03, 20 June 2011 (UTC)

Editors have to accept that there are usually conflicting interpretations of just about anything. That's why 'truth' can't be a criterion for inclusion in WP. The responsible way to handle it is to attempt to present all sides of an issue. Edit wars could be greatly reduced if, instead of reverting or replacing single-sourced, controversial text, editors instead add alternative viewpoints (well-sourced, non-OR and verifiable, of course). Very contentious issues often rate their own sections within an article, or even separate articles. WCCasey (talk) 19:22, 20 June 2011 (UTC)

Deciding which sources are "well-sourced" (and thus likely to be true) just functions to hide the whole issue. You can't decide which sources are good till you know the prevailing scholarly view yourself. There are exceptions, but you usually need to BE a scholar or academic to know this about source-quality. Most Wikipedians are NOT, and thus concensus opinion will dilute out academic and scholarly opinion. There is no fix for this without a change in policy. To put it in a nutshell, you can't demand non-scholars to know what the accepted and best-reputation scholarly sources are. They don't know, but they still have an input to consensus editing, and that degrades your product. SBHarris 19:32, 20 June 2011 (UTC)

It goes even deeper than that as some of those agenda driven pieces are put out by scholars themselves and unfortunately scholar do produce unreliable material (as scholarly reviews of scholarly work show) and WP cannot really fix that or be the referee which scholars got it right. As far as the consensus editing is concerned that isn't necessarily degrading the product, that would be depend on the actually consensus. If consensus results in providing more details on notable different scholarly opinions rather than just one mainstream view it can often be a plus, if it results in including unnotable fringe for consensuss sake than indeed it degrades the quality. Another thing is that in these discussion the primary goal of WP (and its real achievements) often gets overlooked. WP's primary goal is not to provide the optimal in depth scholarly summary on some on some topic, but rather providing some reasonable/sufficient (but far from optimal) summary on a waste variety of topics (for free), i. e. being a free general encyclopedia. The latter is exactly why a policy change will not really help.--Kmhkmh (talk) 20:47, 20 June 2011 (UTC)

• (edit conflict)But that's not what we do, WCCasey. It's what policy says we should do, but in practice, Wikipedia comes down on one side or another, for example on objections to evolution or ancient astronauts. Where there's controversy Wikipedia broadly follows the scientific consensus, and we have a few pages where Wikipedia explicitly debunks published sources on the strength of other published sources (such as list of common misconceptions).

  This is normally rationalised in terms of selecting the most reliable source. But unfortunately, that's where the logic breaks down, because you can't decide which source is the most reliable, credible, trustworthy etc. without deciding which is the most likely to be accurate, and that in turn depends on which is most likely to be true. In other words, our behaviour in the real world of editing is not always about verifiability at all. It's about truth. And that's as it should be. We're an encyclopaedia, so our basic aims are to summarise the topics we cover, to educate, and to inform.

  In practice, most Wikipedians care about the truth and try to tell it.—S Marshall T/C 19:59, 20 June 2011 (UTC)

Hello, interesting discussion but does it translate into actually upholding verifiability in practice? The problem illustrated by 81.103.121.144's concerns - and his/her editing activity on several articles apparently impacting on the Jewish-origin myths of the Syrian Malabar Nasrani community in Kerala - is when religious or ethno-religious "truth" conflicts with, as 81.103.121.144 says, "thousands of published peer reviewed books and papers out there which regurgitate the same old mistakes", i.e. a giant conspiracy of dictionaries, encyclopedias and mainstream scholarly works which in this particular case are surpressing the "truth," in this case of the secret origins of Christianity found in the early medieval Toledot Yeshu, but it could be any "truth" known with great conviction to the IP, and "truth" confirmed by the conspiracy of WP:source and WP:fringe policies against the "truth".

Which is why "verifiability" is a better standard than "truth" ....but "verifiability" is no use when a couple of anon IPs are not bound by WP:3RR while registered editors are, for example. Not the only problem in this case, but an illustration. Please excuse the whinge. In ictu oculi (talk) 20:56, 20 June 2011 (UTC)

Here's what's going through my mind.

“ Verifiability, not truthiness, is a minimum standard for inclusion on Wikipedia. Please make sure that any content has been published by a reliable source before adding it to an article, and please cite your source in such a way that readers who have access to the source can easily check it for themselves. Where the sources disagree, but there is a consensus of academic professionals, please follow the academic consensus. Where there is disagreement among academics as well, please describe the dispute rather than taking sides. Wikipedia mainly publishes mainstream thought, but may mention minority or fringe views. If you think most of the sources are in the wrong, then Wikipedia is not the place to set the record straight. ”

It's not a workable draft because of that "truthiness"; I'll probably need to substitute a whole phrase for that.—S Marshall T/C 21:30, 20 June 2011 (UTC)

There is an endlessly complex array of possible scenarios. The two that seem to be used an an example for everything: opposing views on a clear statement of a matter of fact (did we actually land on the moon) and opposing views on a matter of opinion (is Obama a good or bad President) I think cover a minority of the cases.

Probably the more common case is where some quarter given to accuracy would help the discussion considerably, but the persons who's goals would be furthered by inaccuracy can use wp:policies or lack thereof to exclude any quest for or consideration for accuracy. And so, rather than trying to wade into the complexities or sorting out conflicts, here's a simple 6 word addition which would tip the balance back towards middle ground. That is to add the following radical statement into a prominent and authoritative place: "Accuracy is a goal of Wikipedia."

Sincerely, North8000 (talk) 23:47, 20 June 2011 (UTC)

It is correct that many historical topics are seen from different perspectives in different places; for example, a war between two countries may have undisputed details (battles, dates, commanders, etc.) but different descriptions at both countries in details such as the causes leading to the war, the consequences, the legitimacy of this or that war action, etc. The best way to deal with this is to write sections or articles detailing the historiography of the event, rather than the history of it Cambalachero (talk) 01:24, 21 June 2011 (UTC)

Proposal

This proposal grows out of #A suggestion, above, and is an attempt to distill all of the discussions here. The opening of the page currently reads:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true.

I propose to change it to:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether anyone one can check that material in Wikipedia has already been published by reliable sources, not whether individual editors think it is true. Wikipedia strives to present what those reliable sources say in as accurate a way as possible.

--Tryptofish (talk) 21:38, 18 July 2011 (UTC)

• Support as proposer. There are various wording tweaks, but what I think is important is the new second sentence: it keeps the iconic "verifiability, not truth" in the first sentence, but it goes some way towards addressing the concern that we shouldn't imply that truth doesn't matter. --Tryptofish (talk) 21:42, 18 July 2011 (UTC)

• Request Would you consider amending it it a bit to the one from above: "Verifiability helps ensure accuracy, and Wikipedia strives to present what reliable sources say in as accurate a way as possible"  ? — Preceding unsigned comment added by North8000 (talk • contribs) 23:28, 18 July 2011

It's not what I propose, and I would lean towards WP:KISS, but I'm open to anything. Let's see what other users say. --Tryptofish (talk) 18:46, 19 July 2011 (UTC)

Then I'd have to oppose it as written. North8000 (talk) 21:09, 19 July 2011 (UTC)

As is your right. But, especially per what you said before, perhaps you might not want the perfect to be the enemy of the good. --Tryptofish (talk) 21:37, 19 July 2011 (UTC)

I think I mis-spoke. What I really should have said that without that, this such a tiny change that it it is not the middle ground that will set the issue aside for now. North8000 (talk) 23:50, 19 July 2011 (UTC)

• Oppose. Not an improvement. WCCasey (talk) 03:56, 19 July 2011 (UTC)
• No objection to Tryptofish's version... but not convinced its needed, as this is all covered at WP:NPOV. I don't like North's alternative... I think it is overly wordy and confusing. Blueboar (talk) 21:15, 19 July 2011 (UTC)
• I think we're looking for a short way to say: "Not truth" means "don't include anything unverifiable, even if you think it's true." It doesn't mean "it's okay to add falsehoods so long as someone's printed them." Ideally we might want to add a link to WP:HANDLE which enjoins editors to correct inaccuracies if they find them.—S Marshall T/C 21:37, 19 July 2011 (UTC)

I'm receptive to that. Would it be too much of an Easter egg to link "strives to present" in the second sentence to that? --Tryptofish (talk) 21:41, 19 July 2011 (UTC)

Re: dealing with "Falsehoods"... The problem is that people can disagree over whether something is "False" (which is another way of saying they disagree as to whether it is "True")... sometimes we have to include discussion of things that we may believe to be false, for the simple reason that someone else (ie a reliable source) believes it to be true (this is what WP:Neutral point of view is all about). Not always (WP:UNDUE and WP:OR both give examples of times when we don't)... but often. Perhaps the confusion here stems from the fact that while we have only one criteria for inclusion (Verifiability) we have multiple criteria for exclusion. Blueboar (talk) 22:08, 19 July 2011 (UTC)

Yes, and my thinking is that, here at V, we might best focus on accuracy, accurate presentation of what is verifiable. --Tryptofish (talk) 22:12, 19 July 2011 (UTC)

• Support. I'm not convinced it is needed, either, but enough editors have expressed concerns that I think working to address those makes sense. I'm not opposed tweaking the wording of the second sentence, either. I think it might be helpful if an editor thinks a tweak would move that editor from oppose to support would propose that tweak in the oppose, as that will help us get a sense of direction. --Nuujinn (talk) 22:21, 19 July 2011 (UTC)
• Support for the same reasons as Nuujinn. — V = IR (Talk • Contribs) 22:59, 19 July 2011 (UTC)
• Support Tryptofish's version (North's addition confuses the issue a bit, I think). Explicitly endorsing the concept of "accurately presenting sources" seems helpful (even though that's what WP:NPOV is all about, as Blueboar mentioned). Mlm42 (talk) 23:15, 19 July 2011 (UTC)

Reconsidering, I've struck my support.. A Quest For Knowledge has a point: perhaps this is better left ambiguous. Mlm42 (talk) 19:15, 20 July 2011 (UTC)

• I don't think that the second sentence constitutes an improvement, but I don't really care. However, I strongly object to the "anyone" change in the first sentence. "Anyone" will mean "everyone" to most of our editors. Material is still verifiable even if you personally aren't able to read the source for free. We do not want to have people saying that "it says anyone has to be able to check, and I can't read the source, and so this WP:PAYWALLed is banned by the policy". The word you doubtless meant to use there is someone. WhatamIdoing (talk) 23:37, 19 July 2011 (UTC)

  Now that I think about it, I prefer "readers" over "anyone" as well... but not for the reasons WhatamIdoing gives. I prefer "Readers" because it focuses the statement on our audience... as opposed to "editors". Blueboar (talk) 23:52, 19 July 2011 (UTC)

  Actually, I tend to agree with the "anyone" criticism. Originally it was "someone", which seems more appropriate to me, but tryptofish thought that using it generated a "who?" response. I don't know about "readers" either, since that seems to specifically exclude editors, which seems somewhat odd.
  — V = IR (Talk • Contribs) 00:01, 20 July 2011 (UTC)

• Oppose "Wikipedia strives to present what those reliable sources say in as accurate a way as possible." No, it does not. Wikipedia editors strive to have decent articles and then link to reliable sources to verify the information. We don't exist to summarize what someone else says. No sense adding this, and it could cause some confusion. I agree with WhatamIdoing. Anyone is a horrible addition since all references to books that every single person doesn't own, and newspapers and magazines that archive their older stuff behind paywalls, becomes invalid. Dream Focus 23:54, 19 July 2011 (UTC)

Dream... Re: "We don't exist to summarize what someone else says"... I could not disagree more... Wikipedia is a tertiary source... by definition a tertiary source is one that summarizes what is said in secondary (and to a lesser extent, primary) sources. So, summarizing what others (ie reliable sources) say about a topic is exactly what we exist to do. In fact, we have a policy that explicitly disallows material that isn't a summary of what reliable sources say (see WP:No original research). Blueboar (talk) 00:19, 20 July 2011 (UTC)

I mean, we don't just take one source and copy over everything they say, rewriting it just a bit to avoid copyright problems. Dream Focus 00:28, 20 July 2011 (UTC)

No, we don't and that is addressed in WP:COPYVIO. What in the wording proposed do you object to other than the use of "anyone"? --Nuujinn (talk) 00:35, 20 July 2011 (UTC)

• Comment. Looks like we have two unrelated changes together here. Hate to say it but we should sort that out first. North8000 (talk) 23:57, 19 July 2011 (UTC)

I think you're right, but I'm not sure I see a way around "anyone" versus "someone", as both have a narrow scope definition, and a wider scope in everyday usage. Are there any other ways to indicate that? --Nuujinn (talk) 00:35, 20 July 2011 (UTC)

I was thinking more just about the poll process.....to separate the questions for those two unrelated proposed changes....without that this can only head into a jumble. North8000 (talk) 01:00, 20 July 2011 (UTC)

• Oppose The added phrase, "Wikipedia strives to present what those reliable sources say in as accurate a way as possible" ironically emphasizes verifiability over factual accuracy - which I thought the complaint was about. The new sentence says that editors have to faithfully repeat mistakes in otherwise reliable sources, which I object to. In any case, it ain't broke, don't fix it. A Quest For Knowledge (talk) 01:09, 20 July 2011 (UTC)
• About "anyone": I'm having a facepalm moment! I agree with the criticisms of that word! Somehow, I lost track of this when I made this proposal, but I had intended all along to simply say "one". Changed accordingly. (If there are procedural objections: whatever.) --Tryptofish (talk) 15:23, 20 July 2011 (UTC)

I'm shocked that you're not able to retain the entirety of this short discussion with all wording variants at your fingertips, shocked I say! ;). Seriously, good choice, using a more formal form lifts the phrase from the common usage variations to which both "someone" and "anyone" are subject. --Nuujinn (talk) 15:33, 20 July 2011 (UTC)

Yeah, as I remove the spinach from between my teeth, what I can reconstruct was that I had floated the idea of "one is able to", but Ohms correctly pointed out that it was too formal, to the point of being stilted. It later popped into my mind to say "one can", but apparently it also popped out. --Tryptofish (talk) 15:52, 20 July 2011 (UTC)

• Support With the caveat that, as written, this is such a minor change that it isn't going to resolve the issue. North8000 (talk) 15:45, 20 July 2011 (UTC)
• (re-adding this after 5 edit conflicts, so I haven't read all of the (4?) posts above yet) Here's a new proposal, with a rational: Just leave it as is. Look, different people interpret the important of verifiability to different degrees. To some, verifiability is paramount, while to others accuracy is paramount. Either extreme, taken to an absurd level, is equally damaging. The shades of grey in between aren't incorrect though, they're simply arguable (which is exactly what we've been doing). I don't fault people for feeling that accuracy is more important than faithfully reproducing what the sources say. In the end, all of this is a "feel" thing, and it also relates to considerations of Original Research, Neutral Point of View, and the specifics of the instance in question. I don't think that there's any formulation that we could make here which would appropriately balance every consideration, so I think that we should be silent on the details here in favor of allowing discussion to occur where specific issues come up in articles.
  — V = IR (Talk • Contribs) 15:53, 20 July 2011 (UTC)

For a while there, the servers were acting up; I hope they've settled down now. I actually don't care terribly much, and only recently started watching this talk page. But it seems to me that a lot of editors are unhappy with the status quo, because of the "not truth" issue, while at the same time a lot of editors (including me) are strongly opposed to removing the "not truth" phrase. Thus, a bigger change won't have consensus, but no action will mean no end to debate. --Tryptofish (talk) 16:37, 20 July 2011 (UTC)

Eh... there may be a lengthy debate (hell, there has been a lengthy debate), but it'll end eventually. I'm just trying to make a case for the Status Quo position, which really says just as much about the issue as adding or changing something in the policy. Discussing these things until we can find some sort of agreement doesn't bother me in the least.
— V = IR (Talk • Contribs) 17:12, 20 July 2011 (UTC)

The recently archived RfCs seem to indicate a community that is split over the "not truth" issue. --Tryptofish (talk) 17:21, 20 July 2011 (UTC)

Which is why we're still discussing it. All I'm saying is that "do nothing" is just as much of a choice as "add something!", you know? Actively deciding not to say something can speak volumes.
— V = IR (Talk • Contribs) 17:42, 20 July 2011 (UTC)

Yes, but if we don't come to some kind of compromise by early august, the economy will collapse and we'll fall back into recession....

Seriously, if we can address the concerns of fellow editors, and reach consensus, that's a better outcome, I think. --Nuujinn (talk) 17:46, 20 July 2011 (UTC)

LOL! (actually, I'm somewhat scared of "the economy will collapse", personally...) What I'm offering here though is addressing the concerns of fellow editors, though. the difference is, it's addressing their concerns through means other than adding something to the policy. I'm not saying "I don't want to hear that", but exactly the opposite: "I hear what you're saying. I think that we should address this on a case-by-case basis, rather than through a change to policy." Now, if North chimes in with something about this being a poor solution then I know that I've hit on a true compromise solution. [[=)|grin}}
— V = IR (Talk • Contribs) 18:09, 20 July 2011 (UTC)

• weak support. Both variation are ok, though I consider the new suggestion as marginally better. But as I said at the end of the day both work and a seemingly endless discussion about such marginal difference is a big waste of time.--Kmhkmh (talk) 18:06, 20 July 2011 (UTC)

As I'm starting to learn! --Tryptofish (talk) 19:49, 20 July 2011 (UTC)

Back to my earlier request/suggestion, I think that with the additional 4 words in the new sentence (so it would be "Verifiability helps ensure accuracy, and Wikipedia strives to present what reliable sources say in as accurate a way as possible" un-disses the idea of seeking accuracy, and IMHO would be enough of a change to mostly settle this North8000 (talk) 20:02, 20 July 2011 (UTC)

North, As we have seen, different people can read all sorts of unintended meaning into a simple phrase... so... Could you explain what you mean by "Verifiability helps to ensure accuracy"? It might help to give us an example of how verfifiability (as you understand the concept) does this? Blueboar (talk) 22:14, 20 July 2011 (UTC)

In terms of the structural operative statement of the sentence, it means nothing, just as "not truth" means nothing in the structural, operative statement of that statement. In terms of the psychological impression left, it legitimizes a goal of accuracy, just as "not truth" de-legitimizes it, sort of a balancing statement regarding that. Sincerely, North8000 (talk) 02:02, 21 July 2011 (UTC)

• Oppose, per A Quest for Knowledge. This is well intended but from my point of view what it solves isn't the actual problem. What we're worried about here is rubbish that's appeared in print. We're trying to prevent our policies opening the way for von Däniken or Velikovsky to come into the encyclopaedia. We want a brief wording that explains that creationists and climate change deniers don't get equal time, and why. It needs to be short and pithy.—S Marshall T/C 22:22, 20 July 2011 (UTC)

My reading (please correct me if I'm wrong) of Quest's comment was that it was about not repeating errors from reliable sources, whereas I understand you to be talking more about sources that really are not "reliable". Is it really the case that those kinds of sources pass WP:RS#Reliability in specific contexts for these purposes? And if the issue is "equal time", as opposed to inclusion at all, isn't that already covered by WP:UNDUE? --Tryptofish (talk) 22:35, 20 July 2011 (UTC)

• Yes, part of what I'm talking is about fringe sources, careless sources, and apparently good-faith sources that nevertheless deviate from the mainstream academic consensus. Yes, it's true that all the detail of that should be covered in WP:RS and WP:UNDUE; all that needs to go in WP:V is a short phrase and a signpost to the details held elsewhere. But the short phrase is important.

  The current wording is "The threshold for inclusion in Wikipedia is verifiability, not truth." I don't think we should go into "truth" at all, I think it's an unnecessary can of worms, and I think the ideal phrasing would simply be "the minimum standard for inclusion in Wikipedia is verifiability." But since we're apparently forced to add words to this policy rather than subtract them, we need to explain what we mean by "not truth": (1) that editors don't get to add unsourced unsourceable information even if it's true, and (2) just because something has been published, that doesn't mean it's suitable for inclusion.—S Marshall T/C 22:55, 20 July 2011 (UTC)

On Point 2 - the problem is that something may not be suitable for inclusion in one article, but may be perfectly suitable in another. (for example, It would be absolutely appropriate to discuss what von Däniken says in an article on Theories of alien visitation in ancient times... but not appropriate to discuss it in the article on Mayan civilization. That's exactly why we have WP:DUE. Blueboar (talk) 00:36, 21 July 2011 (UTC)

On Point 1, editors do get to add unsourced information. They don't get to add unsourceABLE information. The two words are not synonymous. WhatamIdoing (talk) 19:53, 22 July 2011 (UTC)

Thank you, WhatamIdoing, for quibbling that point. I've fixed it.—S Marshall T/C 21:15, 22 July 2011 (UTC)

Well, let me float this: how about blue linking the word "accurate" in the second sentence of what I propose, to WP:RS#Reliability in specific contexts? That puts it in line with what you are talking about, and takes it away from what might be construed as slavish transcription of errors and fringe views. --Tryptofish (talk) 23:03, 20 July 2011 (UTC)

For clarity, here is the current draft of what I'm proposing, incorporating the corrections that have come up through this discussion:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether one can check that material in Wikipedia has already been published by reliable sources, not whether individual editors think it is true. Wikipedia strives to present what those reliable sources say in as accurate a way as possible.

--Tryptofish (talk) 23:21, 20 July 2011 (UTC)

FWIW, I think that's fine. I think it's pretty clear that whatever editors think the real problem is, there's no consensus to change the first line, so it seems to be the question is, is this suggested version better or worse than what we have now. I think it's better. --Nuujinn (talk) 01:12, 25 July 2011 (UTC)

As explained recently in detail, there is no consensus to keep the first line unchanged.  Unscintillating (talk) 04:42, 25 July 2011 (UTC)

Maybe, but without consensus for a change, there can be no change, and continuing to argue about the first line after literally months of discussion with no consensus in sight on the "verifiability, not truth" issue is, in my opinion, bordering on disruptive. We should move forward. Some of us are working on wording to address everyone's concerns in the subsequent lines. Care to join us int that effort? --Nuujinn (talk) 09:44, 25 July 2011 (UTC)

I don't think that the most prominent recent attempted-middle-ground proposals address the concerns of the folks that want to change the first sentence, so I don't think that they could settle it. I think that the true middle ground would be a second sentence that really covers/clarifies the items of concern. North8000 (talk) 11:54, 25 July 2011 (UTC)

I thought that a second sentence that clarifies the concerns was the one thing we had been able to agree on. Isn't it? I thought the purpose of this discussion was to outline the new sentence's content, after which we could work on detailed phrasing. At the moment, what's on the table is that the proposed second sentence should look at:-

1) That editors don't get to add unsourceable information even if it's true; and

2) That just because something's been published doesn't necessarily make it suitable for inclusion (and it's agreed that this second limb should contain a bluelink to WP:UNDUE).

Are there other items of concern?—S Marshall T/C 12:03, 25 July 2011 (UTC)

First, to get the straw men out of the way, let's stipulate that everybody wants verifiability to be a requirement for inclusion. Also,that where there are NPOV concerns, that some mandated inclusions per wp:npov are in order. That said, a main concern is that the "not truth" in the first sentence gets widely mis-interpreted that accuracy, falseness of a statement, plausibility / implausibility of a statement have NO place. ZERO place in a discussion or editor judgment about what does or doesn't go in the article. Also that something being sourced is a sufficient criteria for someone to force it to be in the article. And so I think that a compromise 2nd sentence should address that. Adding something like "Verifiability helps achieve accuracy" would help in that respect. North8000 (talk) 12:23, 25 July 2011 (UTC)

"Verifiability helps achieve accuracy" is too terse... it is open to misunderstanding. Perhaps if you explained how "Verifiability helps achieve accuracy" we could reach agreement. Blueboar (talk) 14:55, 25 July 2011 (UTC)

I've thought about the discussion above and below, and I have an idea about a way to use a blue link to WP:UNDUE that might help:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether one can check that material in Wikipedia has already been published by reliable sources, not whether individual editors think it is true. Wikipedia strives to present what those reliable sources, taken as a whole, say, in as accurate a way as possible.

(What I changed was the addition of "taken as a whole".) I think (hope?) that this keeps the intended and agreed-upon meaning that you can't add something unsourced because you think it's true, but it also rules out simply saying that you've accurately quoted a fringe or erroneous source. --Tryptofish (talk) 18:48, 25 July 2011 (UTC)

A new suggestion (not a proposal yet)

What about adding something along the lines of:

• However, Verifiability on its own does not guarantee or mandate inclusion. Wikipedia has many other policies and guidelines that can impact how, when, and even whether some specific piece of information is actually included in a specific article (especially, but not limited to, our WP:Neutral point of view and WP:No original research policies). Verifiability means we can consider inclusion, but does not mean we must include it.

For now... think about the concept of what I am suggesting, not the exact wording... I think something like this would address the concerns raised above... making it clear that Verifiability is not the be-all-and-end-all of inclusion... merely the first necessary step in a complicated process. It does not even have to be the second sentence (although I would include it in the lede). Blueboar (talk) 13:23, 25 July 2011 (UTC)

• Whatever we include needs to be as brief and terse as possible. I think we've agreed on one sentence. I'd certainly be opposed to adding a whole paragraph of explanation when the problem is only caused by two words ("not truth") that aren't essential to the policy's meaning in the first place.—S Marshall T/C 13:43, 25 July 2011 (UTC)

I don't think we have agreed on much of anything... except to consider adding a second sentence (if we can agree on wording). And we have especially not agreed that the two words "not truth" are "not essential to the policy's meaning". (Indeed, many of us think those two words are not only essential but vital). As a general idea, I can agree that brief and terse is good... but perhaps the problem is that we are trying to be too brief and terse. Blueboar (talk) 14:04, 25 July 2011 (UTC)

How 'bout: "Verifiability helps achieve accuracy". North8000 (talk) 14:15, 25 July 2011 (UTC)

(answered above... let's keep discussion of different suggestions to different sections) Blueboar (talk) 14:58, 25 July 2011 (UTC)

• (going back to the topic) Irrespective of whether "not truth" is vital, certainly we can't add a whole paragraph. I think we want to focus on the point that articles should be both verifiable from published sources and factually correct. (I also think this "factually correct" point is where North8000 is going with his "accuracy" phrasing). How about:-

  "The threshold for inclusion on Wikipedia is verifiability, not truth--whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Verifiability is necessary but not sufficient for inclusion--in other words, a source is needed, but even if a source exists, its content is not necessarily suitable for the encyclopaedia--and the phrase 'not truth' does not make it acceptable to introduce inaccuracies into articles."

  Disclaimer: For the purposes of this discussion I've held my nose and included the words "threshold" and "not truth" in my proposals. This should not be taken to mean that I have changed my position on these phrases. I haven't.

  I know that wording needs tweaking and simplifying. Elegance escapes me today. Suggestions are welcome.—S Marshall T/C 18:48, 25 July 2011 (UTC)

Re: "and the phrase 'not truth' does not make it acceptable to introduce inaccuracies into articles" I have a problem with this ... it could be used to inappropriately exclude POVs that should be included (on the grounds that they are "inaccurate"). I agree that we should be able to correct some kinds of "inaccuracies" (like obvious typos) ... but the fact is, some "inaccuracies" are actually POVs that need to be included. Blueboar (talk) 19:22, 25 July 2011 (UTC)

Taking the compromise of leaving the first sentence as-is as a given, I like it. North8000 (talk) 18:51, 25 July 2011 (UTC)

I wonder if a workable approach would be to say: "Verifiability is necessary but not sufficient for inclusion.", and then have the revised version of the last sentence that I just suggested above. --Tryptofish (talk) 19:14, 25 July 2011 (UTC)

Or maybe in the opposite order? --Tryptofish (talk) 19:15, 25 July 2011 (UTC)

• @Blueboar:- I was hoping to use "inaccuracies" as a shortcut for "views contrary to the mainstream academic consensus". What I'm getting at here is that we should present the opinion held by the majority of scientists or academics working in the relevant field as if it were fact. If there are two or more mainstream opinions then we present the dispute without taking sides. In other words, I'm trying to document what we do in cases like global warming: there are two views, one held by the vast majority of scientists and one held by a small number of politicians and wealthy special interests, so we present the mainstream scientific view ("accurate") as if it were true and we keep the fringe view ("inaccurate") at a distance. If I unpacked the meaning I'm trying to get into "inaccuracies" then the whole sentence would read:

  "Verifiability is necessary but not sufficient for inclusion--in other words, a source is needed, but even if a source exists, its content is not necessarily suitable for the encyclopaedia--and the phrase 'not truth' does not make it acceptable to introduce any views contrary to the mainstream academic consensus into articles except in accordance with WP:UNDUE."

  But I hate that, for obvious reasons of length. Do you agree with what I'm trying to say, even if you don't agree with the word "inaccuracies"?

  @Tryptofish:- I really think we need to nail down this "inaccuracies" point. I like the "taken as a whole" wording, but I'm not yet sure if or how to include it. Can we put it in the "good idea" box for the moment and keep brainstorming?

  @North8000:- We might have to lose the "accuracy"/"inaccuracy" wording, because let's face it, Blueboar has a point. I think this discussion shows that an inescapable consequence of the "not truth" part of this is that we have to define "truth" before we can get down to cases. I propose to do that by reference to the mainstream academic view.—S Marshall T/C 20:22, 25 July 2011 (UTC)

I was thinking of a vagary to offset a vagary, a non-operative phrase to off set a non-operative phrase. Saying the word "accuracy" in a positive light to offset the denigration of the accuracy quest done by "not truth". I agree / think that if we try to go much further than that with the word "accuracy" it will turn into a quagmire. North8000 (talk) 20:58, 25 July 2011 (UTC)

S Marshal - I can fully accept the concept behind "we should present the opinion held by the majority of scientists or academics working in the relevant field as if it were fact" (and the "two or more mainstream opinions" qualification), however (and this is a huge however) this does not mean we can exclude the presentation of the opinions held by the minority of scientists and by non-academics... and that is my concern with your language. I think "mainstream POV" pushers will misconstrue your proposed language to mean that only the mainstream view should be included. NPOV makes it clear that we often must include ("inaccurate") minority and non-accademic views (per WP:DUE)... the difference being that we can present such views as being "opinion". Blueboar (talk) 21:07, 25 July 2011 (UTC)

Hmmm... this may be a fruitful distinction... raw inclusion/exclusion determination as distinct from presentation as fact/opinion. Blueboar (talk) 21:19, 25 July 2011 (UTC)

• I have a question. What's really wrong with adding a paragraph (or, at least, fleshing out the hanging first sentence into a full paragraph)? There's certainly no need for unrestricted verbosity, but we don't need to be terse to the point of causing misunderstanding (which is where I think we currently are, based on... all of this).
  — V = IR (Talk • Contribs) 21:18, 25 July 2011 (UTC)
• @Blueboar:- I was inching towards that with we present the mainstream scientific view ("accurate") as if it were true and we keep the fringe view ("inaccurate") at a distance. It's grounded in the examples we've already discussed (Apollo program vs. Moon landing conspiracy theories, etc.) There can be several views, but the mainstream academic view shall be in the simple indicative ("Neil Armstrong landed on the moon") and the alternatives, if there is a consensus to express them at all, shall be in the reportative ("Source X disputes that Neil Armstrong landed on the moon"). But it's a bit tangential to the main point.

  @Ohms law:- The problem I see is that editors whose attention is directed to this policy may not have a very long attention span. I'm anxious to try to keep it within the TLDR safe-zone. I wouldn't agree to a whole new paragraph unless convinced it was necessary and I'm a long way from that!—S Marshall T/C 21:55, 25 July 2011 (UTC)

  I agree that hitting a TLDR point is bad (which is the point that I was making re: being overly verbose), but I don't think that we're anywhere near that point. As a matter of fact, this sprawling discussion is exhibit #1 to the case that we're being overly terse here, to me. Besides, I don't think that policy pages are really intended to be consumed in the same manner as articles are. We can be as verbose as needed to be understandable, as long as we organize the page(s) as a whole into bite sized subsections (including through paragraphs).
  — V = IR (Talk • Contribs) 22:06, 25 July 2011 (UTC)

S Marshall I laud your attempts to get deeper into / more specific regarding accuracy, but think it could turn into a quagmire because you are dealing with just one of the many cases. I'll give you a pervasive but common one. An otherwise-reliable source (and wp:rs) makes an error and uses a clearly erroneous (negative) adjective to describe Politician X. It is so clearly erroneous (like saying Obama has three eyes) that it has never been said elsewhere and so there are no other reliable sources that say otherwise (like no sources that say he has two eyes). Also, so clearly false that none of the combatants will discuss whether it is false or not, and say that (per "not truth") policy says that accuracy is an illegitimate conversation. The persons who want politician X to sound bad want to keep the word in. They say "accuracy" has no place in the conversation, that a wp:"RS" said it, and nobody can produce a RS that directly says the opposite, so it stays in. I would like a few words in the policy that would vaguely tend to legitimize accuracy as a valid consideration in the conversation. For me this is a real world, common representative example to deal with. North8000 (talk) 22:36, 25 July 2011 (UTC)

S Marshall, my answer to the question you asked me way above is: yes, of course. But, that said, I'm having trouble thinking of anything having to do with "inaccuracies", as they are covered by WP:UNDUE, that would not be covered by taking the reliable sources "as a whole". It seems to me that everything that you raise here as "inaccuracy" is what comes from either cherry picking sources, or from giving some sources undue weight. Do I misunderstand? --Tryptofish (talk) 22:53, 25 July 2011 (UTC)

• @North8000:- You probably already know that I'm basically with you on this.  :) We've got to define what "accuracy" means before we'll get anywhere with this, I think. I want to avoid being circular by defining "accuracy" as either "uncontroversial fact" (e.g. Obama has two eyes, or Paris is the capital of France) or else "mainstream academic viewpoint" (e.g. species exist because of evolution and baraminology is bunk). The alternative is to define accuracy in terms of what is and isn't true, which is a major problem in a policy that begins with "not truth". Yes, I know it's awkward and unnecessary.

  @Tryptofish:- Cherry-picking sources is what encyclopaedia editors are supposed to do. We're supposed to evaluate the sources, select the best ones and summarise what they say. We're not required to indiscriminately use all the sources available and I don't think that's even a good idea. I agree wholeheartedly on the weight point, and I like the idea of linking "taken together" to WP:UNDUE. I don't yet know how to include it in this sentence that we're considering, or even if this sentence is the best place to put it; I hope other editors will chime in on this.—S Marshall T/C 23:37, 25 July 2011 (UTC)

What we are talking about is summarizing how WP:NPOV (and especially WP:DUE) interacts with WP:V. I don't think that can be done in one sentence. Blueboar (talk) 23:47, 25 July 2011 (UTC)

Blueboar, wp:undue kicks in when there is a NPOV question. One of it's biggest flaws (that it only covers items that reflect on a question-in-dispute) even further narrows that. I think that we should leave it to wp:npov to cover those cases. One example is being selective about ("cherry picking") sources. This is an essential process for normal good editing, but needs to stopped when there is an npov type dispute.

If you are reflecting on my specific example, at the granular level it is also a wp:ver wording problem. The "not truth" wording allowed folks to say that accuracy (clear falseness of the item) could not even be a consideration in the conversation.

S Marshall, maybe you are right, this could be simpler than I thought when you spoke of defining accuracy as uncontroversial fact, then that means that accuracy can be taken into consideration at least until it is decided that it is controversial. North8000 (talk) 10:06, 26 July 2011 (UTC)

Re: The "not truth" wording allowed folks to say that accuracy (clear falseness of the item) could not even be a consideration in the conversation. .... I have to ask: which conversation? In a conversation that is about verifiability (whether some potentially inaccurate factoid is verifiable or not), I would agree that discussions about accuracy should not be part of the conversation (the source either supports the item or it does not). However, in a discussion about the reliability of the source, or the proper neutral presentation of the item (especially due weight discussions), then accuracy can and should be part of the conversation.

When a POV pusher argues: "I don't care if you think X is inaccurate... WP: V says "Verifiability, not Truth"... The source says X... X is verifiable... it stays in the article", I think the proper reply (assuming the source does indeed say X) is usually: "Verifiability is not the be-all-and-end-all of inclusion... it is only the first step...there are other policies and guidelines that impact inclusion". In other words don't argue about Verifiability... concede Verifiability... and shift the conversation to discussing other policies and guidelines. Focus the conversation on WP:RS (ie question whether the cited source is reliable for that specific fact) or WP:NPOV (ie discuss presentation and Due Weight). Once you have changed which policy/guideline you are talking about... then questions of accuracy are appropriately part of the conversation again. Blueboar (talk) 12:38, 26 July 2011 (UTC)

Been there, done that. The end result was that by doing all of that, and people spending about 10,000 words of discussion, (a very good guess, not a wild exaggeration) the erroneous item got removed. ~10,000 words needed to fix that is a BIG problem. The response through the whole 10,000 words was per the above, and things like "our job is not to decide what is true, it's to report what RS's say". Or, "find a wp:RS that says otherwise". Or, "any statement to the contrary is OR". BTW, can you find a wp:RS that says that OBama is not a right wing extremist or that he does not have three eyes? If not, either of those statemetn would be OR. I only brought up that example and am taking it further because I think it is a good illustration of several common phenomenon, how the wording has promulgated accepted chants which are misquotes of wp:ver, and how it allows allows people to powerfully and almost-invincibly mis-use wp;ver when it serves their purposes. North8000 (talk) 14:12, 26 July 2011 (UTC)

To which I would reply that there is no need for me to find a source that Obama is not a right wing extremist or that he does not have three eyes... since I have no intention of including a statement actually saying that he is not a right wing extremist or that he does not have three eyes. We don't require Verifiability for things that we don't say in an article, only for things we do say. So... we are only talking about whether to include the affirmative statements that he is a right wing extremist, and that he does have three eyes.

If you have already laid out multiple arguments, pointing to multiple policies and guidelines... all indicating that the statement in question should not be included... then the situation has gone beyond a simple policy or content dispute... the other editor's continued insistence on inclusion becomes a question of "WP:IDIDN'THEARTHAT", and thus disruptive behavior. It is time to bring in an admin to deal with that disruptive behavior. Blueboar (talk) 15:27, 26 July 2011 (UTC)

• Could we possibly return to discussing the phrasing of the proposed sentence?—S Marshall T/C 16:03, 26 July 2011 (UTC)

Sure... I will now formally propose:

• The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. However, passing the threshold of Verifiability does not, on its own, guarantee or mandate inclusion. Wikipedia has many other policies and guidelines that can impact how, when, and even whether information is included.

OK... it's not one sentence, but I think it is still terse and to the point. Blueboar (talk) 16:28, 26 July 2011 (UTC)

Okay, starting from Blueboar's revised lede with the three-sentence structure, and tweaking a bit to include the views expressed by North8000, Tryptofish and myself, I get:-

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Verifiability is necessary but not sufficient for inclusion—other policies and guidelines can impact how, and even whether, information is included. Despite the words "not truth", content should still be accurate, and present what the sources taken as a whole say; where sources conflict, the mainstream academic view should prevail. ”

Progress?—S Marshall T/C 17:36, 26 July 2011 (UTC)

I can fully agree with the part of this that is a re-write of my language... I can not agree with the addition (the part that starts with: "Despite the words....".) I have serious concerns (and even strongly disagree with some phrases) in your addition. Your addition could easily be misused to omit minority viewpoints (and even some notable Fringe viewpoints) that WP:NPOV says must be discussed. I strongly disagree with "content should still be accurate"... no, Content does not need to be "accurate" to be included... content needs to be presented acurately when/if included. I strongly disagree with "where sources conflict, the mainstream academic view should prevail"... No... the mainstream view should be neutrally presented as being such, and the minority views should be neutrally presented as being such... but one does not "prevail" over the other. Blueboar (talk) 18:11, 26 July 2011 (UTC)

Impact (IPA: ˈɪmpækt) vt.: What a rock does when it intersects with your head at high velocity.

The word you want is either influence or affect. WhatamIdoing (talk) 18:06, 26 July 2011 (UTC)

Good point... of the two, I prefer "affect" (unless there is a third word that is even better). Blueboar (talk) 18:18, 26 July 2011 (UTC)

"Verifiability is necessary but not sufficient for inclusion." I'm fine with that.

"However, passing the threshold of Verifiability does not, on its own, guarantee or mandate inclusion." Seems no different in meaning than the previous sentence, and is much, much wordier.

"Wikipedia has many other policies and guidelines that can impact how, when, and even whether information is included." (or "—other policies and guidelines can impact how, and even whether, information is included.) A lot of words to say rather little (regardless of impact/affect). How about: Verifiability is necessary but not sufficient for inclusion, because other policies apply along with this one.?

When I used the phrase "cherry picking" of sources, I was, of course, referring to the common use of the phrase to denote POV source selection, not editorial good judgment. I'm still waiting for anyone to explain how "Wikipedia strives to present what those reliable sources, taken as a whole, say, in as accurate a way as possible." fails to cover it. --Tryptofish (talk) 18:27, 26 July 2011 (UTC)

• @Blueboar:- Please start to compromise, because I feel as if I'm making substantial concessions and you're not making any at all. As a starting point, will you please identify any phrasing that you would accept—any phrasing at all—that enjoins editors not to introduce inaccuracies into the encyclopaedia.—S Marshall T/C 19:06, 26 July 2011 (UTC)

I would accept: "See our WP:NPOV Policy for guidance on when you must include material you believe is inaccurate, and when you may omit it."

I can't go beyond that... because WP:NPOV indicates that sometimes you have to include things that you might think are inaccuracies in the encyclopedia, and sometimes you don't. It is a complicated issue, one that takes an entire Policy page to explain. It simply can not be summed up in a one or two sentence "sound bite" here on this page. Blueboar (talk) 21:26, 26 July 2011 (UTC)

But dealing with this is a consequence of including "not truth" in the policy: we have to explain that the phrase "not truth" isn't a licence to lie. So much simpler to omit it. Anyway. Here's proposed wording #4:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Strive to present what those reliable sources, taken as a whole, say, in as accurate a way as possible, but see our policy on neutral point of view for when you must include material you believe is inaccurate. Verifiability is necessary but not sufficient for inclusion; other policies may also apply. ”

Needs to be shorter, of course, but are we getting there now?—S Marshall T/C 22:16, 26 July 2011 (UTC)

About half of everybody wanted to change the first sentence. IMHO, we need to remember that leaving the first sentence as-is, and adding a a second sentence which significantly addresses the issue is the middle-of-the-road compromise. Sincerely, North8000 (talk) 22:25, 26 July 2011 (UTC)

I could support S Marshall's latest proposal... I would prefer that "...in as accurate a way as possible, " be worded "... as accurately as possible", but I definitely don't insist on that change. As for length... why does it need to be any sorter? I think this sums up a complicated issue well. Blueboar (talk) 23:59, 26 July 2011 (UTC)

Tweaked for brevity:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Summarise what those reliable sources, taken as a whole, say, as accurately as possible, but see our policy on neutral point of view for when you must include material you believe is inaccurate. Verifiability is necessary but not sufficient for inclusion; other policies may also apply. ”

It strikes me that the first sentence shouldn't begin with "The threshold", because WP:V isn't the only threshold. I'd prefer "A threshold", or ideally "A minimum standard".—S Marshall T/C 00:09, 27 July 2011 (UTC)

(edit conflict) How about:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Wikipedia strives to present what those reliable sources, taken as a whole, say, as accurately as possible. Verifiability is necessary but not sufficient for inclusion; other policies such as neutral point of view also apply. ”

That says the same things, but is an attempt to make it more succinct. Personally, I wouldn't change "The" at the beginning. --Tryptofish (talk) 00:14, 27 July 2011 (UTC)

Looks good, it incorporates Blueboar's suggestion which I think was a good one. Sincerely, North8000 (talk) 00:49, 27 July 2011 (UTC)

Fine with me. My goodness... does this mean (dare I say it)... consensus? (or at least a consensus amongst the three or four of us who were pig headed enough not to give up on the discussion)? Blueboar (talk) 01:39, 27 July 2011 (UTC)

That is subtle, there is hidden meaning in the Wikilink, whereas with a straight reading, editors may claim an entitlement to the accurate reporting of falsehoods.  A partial fix is to include, along with "other policies such as WP:NPOV", a mention of WP:Editing policy (the policy that states, "on Wikipedia a lack of information is better than misleading or false information").  Also, where do we get in the point that material from reliable sources known to be inaccurate should not use Wikipedia's voice, but rather should use inline attribution of the source?  Unscintillating (talk) 02:47, 27 July 2011 (UTC)

Well, we can't build Rome in this 'lil ole' sentence.  :-) North8000 (talk) 03:12, 27 July 2011 (UTC)

We can try to address what Unscintillating says, though. In particular, that editing policy link strikes me as highly relevant to the paragraph we're trying to concoct.

“ A minimum standard for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Summarise what those reliable sources, taken as a whole, say, as accurately as possible, and no information at all is better than false or misleading information. Verifiability is necessary but not sufficient for inclusion; other policies such as neutral point of view also apply. ”

How's that?—S Marshall T/C 08:27, 27 July 2011 (UTC)

@S Marshal -I thought the goal was to not change the first sentence... try again. Blueboar (talk) 12:33, 27 July 2011 (UTC)

@Unscintillating - re: "Also, where do we get in the point that material from reliable sources known to be inaccurate should not use Wikipedia's voice, but rather should use inline attribution of the source?"... that is done through linking to WP:RS#Reliability in specific contexts (which is the link attached to the word "Accurately" in the text). Blueboar (talk) 12:53, 27 July 2011 (UTC)

Sigh. Here's the first sentence restored in all its glorious wonderfulness, but with the other agreed fixes and clarifications included.

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Summarise what those reliable sources, taken as a whole, say, as accurately as possible, and no information at all is better than false or misleading information. Verifiability is necessary but not sufficient for inclusion; other policies such as neutral point of view also apply. ”

There.—S Marshall T/C 14:15, 27 July 2011 (UTC)

A different approach

Looking at Wikipedia:Editing policy#Adding information to Wikipedia I am a bit concerned that the "misleading or false information" quote is taken out of context ... I have a new suggestion that puts that context back... What about:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Please show that information is verifiable and not original research by referencing reliable sources. Unsourced information may be challenged and removed, because on Wikipedia a lack of information is better than misleading or false information—Wikipedia's reputation as a trusted encyclopedia depends on the information in articles being verifiable and reliable. Verifiability is necessary but not sufficient for inclusion; other policies such as neutral point of view also apply. ”

Does this work for people? Blueboar (talk) 15:09, 27 July 2011 (UTC)

IMHO that sort of goes farther away from addressing the main concern. North8000 (talk) 16:20, 27 July 2011 (UTC)

That's because Blueboar's version strips out Tryptofish's wording, I think, North8000. I think the phrase "Please show that information is verifiable and not original research by referencing reliable sources" is redundant with the second paragraph, which begins "To show that it is not original research, all material added to articles must be attributable to a reliable, published source". The phrase "unsourced information may be challenged and removed" is redundant with the third paragraph, which contains the words "Anything that requires but lacks a source may be removed", and the phrase "Wikipedia's reputation as a trusted encyclopaedia depends on the information in articles being verifiable and reliable" strikes me as unnecessary. If I strip out those three from Blueboar's version, and if I also add back in Tryptofish's preferred wording which the previous version removed, I get:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Summarise what the reliable sources, taken as a whole, say, as accurately as possible, but bear in mind that a lack of information is better than misleading or false information. Verifiability is necessary but not sufficient for inclusion; other policies such as neutral point of view also apply. ”

—S Marshall T/C 16:32, 27 July 2011 (UTC)

Suggest saying that other considerations rather than just other policies apply. North8000 (talk) 17:00, 27 July 2011 (UTC)

I agree that much of what I suggested was redundant... but it is actually a direct quote from WP:Editing policy. It is the context in which the statement "a lack of information is better than misleading or false information" comes from. To quote that phrase away from its surrounding context, in my mind is just as bad as taking "Verifiability, not truth" out of its surrounding context. We should not try to correct one out of context policy misuse by creating another. Blueboar (talk) 17:46, 27 July 2011 (UTC)

• It's not out of its context, though; the context appears in the subsequent two paragraphs.—S Marshall T/C 17:56, 27 July 2011 (UTC)

In WP:EP, the phrase "lack of information is better than misleading or false information" is made in the context of excluding unsourced (and potentially unsourceable) information... but in your proposal you link it to the concept of excluding verifiable (ie sourceable) information. That is taking the phrase out of context. Blueboar (talk) 18:12, 27 July 2011 (UTC)

Okay, no problem, we can have this conversation again if you like. Editors can have reasons to want to insert sourced but false information, such as here. It's ridiculous that the policy on verifiability doesn't just let them do this, it positively disempowers editors who want to prevent it. To verify something means to confirm that it is true. (See wikt:verify). Therefore the whole purpose of these edits is to mitigate the phrase "not truth" in such a way that we say it is not acceptable to introduce untruths into the encyclopaedia. Making the change you propose nullifies the whole point of what we're trying to do.—S Marshall T/C 18:41, 27 July 2011 (UTC)

Due to RL, I can only make a quick comment right now, but how about this: If I understand right, the issue raised by Unscintillating is a matter of an Easter egg link (something I had been uncomfortable about, myself, in fact). Maybe there's a way of going back to that near-consensus version, but unlinking what was disputed, and instead, adding that link, under its actual name, after NPOV as other policies that apply. --Tryptofish (talk) 19:48, 27 July 2011 (UTC)

S Marshall summarized the core of it perfectly: the following is 80% stolen from them: Editors can have reasons to want to insert sourced but false information...the policy on verifiability doesn't just let them do this, it positively disempowers editors who want to prevent it....therefore the whole purpose of these discussed changes is to mitigate the phrase "not truth" in such a way that we say it is not acceptable to introduce untruths into the encyclopaedia. Or, at least to stop actively dis-empowering editors that want to prevent it. (Only) If we can keep that in mind, I think we can work up something that will actually settle this. Sincerely, North8000 (talk) 22:28, 27 July 2011 (UTC)

In this context, the "lack of information" quotation is going to be interpreted as "I should delete anything that I personally disagree with, because telling people that scientists believe AGW is going to 'mislead' some readers into believing in climate change."

This proposed change is not okay, and there is no way to re-phrase it that will make it be okay. When to exclude verifiable material is not the job of this policy. WhatamIdoing (talk) 23:15, 27 July 2011 (UTC)

I'd be delighted to take this policy back to what its actual job is, and propose that we do that by amending the whole first paragraph to read: A minimum standard for inclusion in Wikipedia is verifiability—whether readers can check that material in Wikipedia has already been published by a reliable source. All the extra verbiage is caused by the confusing and misleading phrase "not truth" which we originally sought to remove, but it's a lot simpler to take it out. I've only put the current phrasing in as a good faith attempt at compromise.—S Marshall T/C 23:35, 27 July 2011 (UTC)

Yes, leaving "not truth" in when there was no consensus to keep it (or to take it out) is a big compromise. Even a second sentence which significantly addresses the issue would not be an equal one, but in the spirit of compromise, I think many would settle for it. North8000 (talk) 23:41, 27 July 2011 (UTC)

Another version

OK, after looking more carefully at the issues raised, I suggest:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Wikipedia strives to present what those reliable sources, taken as a whole, say, as accurately as possible. Verifiability is necessary but not sufficient for inclusion; other policies such as neutral point of view, proper utilization of sources, and the requirement not to add misleading or false information also apply. ”

Does that cover everything? (That said, I'm not really sold on the need to add the part about misleading/false, because I think it ends up stepping on the "not truth" part. I can just hear someone saying "Well, do you want truth or not?!" At the same time, I accept that there are concerns about leaving "not truth" insufficiently explained, and I'm ready to go with consensus about this, either way.) --Tryptofish (talk) 00:18, 28 July 2011 (UTC)

Sounds good. North8000 (talk) 01:31, 28 July 2011 (UTC)

Again, I object to the mischaracterization of what it being said at WP:EP. It is not a "requirement not to add misleading of false information", it is a restatement of WP:V's requirement not to add unverifiable information. So, no I don't agree to this last version. Blueboar (talk) 02:07, 28 July 2011 (UTC)

I'm with Blueboar, for the same reasons. WhatamIdoing (talk) 04:02, 28 July 2011 (UTC)

As I said in my parenthetical comment, it makes me uncomfortable at the least. On thinking further, I object to it too! But I think that's the only problem. --Tryptofish (talk) 21:09, 28 July 2011 (UTC)

We came close to consensus

I think we should go back to the version before we tried quoting WP:EP ... as being the closest we have come to a consensus:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Wikipedia strives to present what those reliable sources, taken as a whole, say, as accurately as possible. Verifiability is necessary but not sufficient for inclusion; other policies such as neutral point of view also apply. ”

Does anyone actually object to this version? Blueboar (talk) 02:17, 28 July 2011 (UTC)

No specific objection, but it does not address the issue and so would not resolve it. North8000 (talk) 02:36, 28 July 2011 (UTC)

I don't object to this, but I begin to believe that I will object to anything that "addresses the issue". I begin to believe that "the issue" is that some editors want the policy to say something that this policy has never said and that I believe that the policy absolutely must not say. Inclusion must never be determined on the basis of my personal beliefs about whether the material in question is "true" or "accurate" or "not misleading" or "not false" or anything else that you want to call it. WhatamIdoing (talk) 04:05, 28 July 2011 (UTC)

That's probably why we talk about "the" requirement for inclusion; once material is eligible for inclusion, in other policies we decide the WP:DUE weight for things that are false, etc.  Unscintillating (talk) 07:33, 28 July 2011 (UTC)

WhatamIdoing, earlier in this discussion we cited quite a lot of examples of where inclusion is determined on the basis of editors' beliefs about whether the material in question is true. For example, pyramids contains no mention of ancient astronauts. Apollo program contains no mention of the moon landing conspiracy theories. Earth describes our planet as an oblate spheroid and contains no mention of any possibility that it might be flat. I could go on and on.

In fact, I think I will. Paul Revere doesn't contain any of Sarah Palin's theories about him. History of Australia doesn't mention the dreamtime. Fossil contains no mention of the possibility that all the fossils were implanted by God in 4004BC.

I submit that in controversial encyclopaedia areas, inclusion is usually determined by what editors believe is true; I submit that this is quite normal practice already, and the only stage that remains is for policy to document practice.—S Marshall T/C 08:31, 28 July 2011 (UTC)

Excellent points, S Marshall, thank you.  Unscintillating (talk) 09:00, 28 July 2011 (UTC)

Why? A flat earth section in Earth would be fine with me, so long as it was sourced, and we have a Flat_earth article. We have Dreamtime, Ancient_astronauts, Creationism, Moon landing conspiracy theories and one could add Palin's view of Paul Revere either in her article or in a "popular culture" section in Revere's article. The only thing V does and the only thing is should do is specify that the material must be verifiable. Truth has nothing to do with it, esp. when editors disagree. We use RS policy to make determinations of which sources are more appropriate for a given topic, and NPOV to settle issues of whether to include a sourceable assertion. Community consensus informed by core policies is how we decide, and I'm sure we sometime get it wrong. Every topic S Marshall mentioned is suitable for inclusion if it is not already included, and V does not determine where something appears, how it is presented, or what sources we value most for a topic area. And in my experience, the worst of the POV pushers are always driven by what they say is the truth, see esp. the quote from Swift attached to point 72 here. --Nuujinn (talk) 11:36, 28 July 2011 (UTC)

Nonsense. This is far, far wide. None of these things are even remotely suitable for inclusion in the articles I mentioned. Can you imagine the Encyclopaedia Britannica vacillating or wavering over the shape of the Earth? Or discussing baraminology as a serious alternative to evolution? Of course they wouldn't.—S Marshall T/C 13:26, 28 July 2011 (UTC)

I think S Marshal is off base here... I don't think it is correct to say that these are examples of "situations where inclusion is determined on the basis of editors' beliefs about whether the material in question is true". For every one of the examples, the reason the articles shouldn't mention them has nothing to do with truth/untruth or accuracy/inaccuracy of the theory... the reason these articles don't mention the theory is that, within the context of the articles in question, mentioning it would give UNDUE weight to a Fringe (or at lest extreme minority) viewpoint. In other words, While WP:Verifiability justifies including these theories somewhere in Wikipedia, WP:NPOV (and especially WP:UNDUE) justifies saying "sure, but not in the context of this article. Blueboar (talk) 11:53, 28 July 2011 (UTC)

I'm quite sure that the vast majority of Wikipedia editors would tell you that the fringe theories I mentioned are wrong. As in, false. Untrue. Misconceptions. Hoaxes. Lies. And I'm sure that attempts to introduce these fringe theories will be removed. That diff is informative: it describes the Sarah Palin interview as "not a reliable source", because the editor removing it has made the value judgment that Sarah Palin was wrong. And he was right to make that judgment.

This business of being open-minded and open to including whatever can be sourced is bad and wrong. An encyclopaedia's job is to inform—to educate. Anyone who's prepared to publish known error in an encyclopaedia, except to debunk it, needs to go and find another hobby.—S Marshall T/C 13:26, 28 July 2011 (UTC)

And I would agree with the majority who say that these fringe theories are "false, untrue, misconceptions, hoaxes, lies." However, my personal views (or those of other editors) do not mean I can or should exclude all mention of them from Wikipedia. We might (and indeed do) exclude them from a specific article (on WP:UNDUE grounds for example), and yet we might (and often do) include them in another article. You are correct in saying "An encyclopaedia's job is to inform—to educate."... and that includes informing-educating people about what notable Fringe theories say. We can (and frequently do) debate where, when and how we do this... and have leway as to whether to do so in a specific article... but that does not mean we can completely exclude them. That is supported by multiple policies and guidelines. Blueboar (talk) 14:10, 28 July 2011 (UTC)

That's just it: The astronaut fairy tale about the pyramids is IMO "wrong", "false", "stupid", etc., but the fact that this story exists is verifiable, which is the sole subject of this policy. The fact that this story exists should not be mentioned in, say, Myocardial infarction (where 100% of editors will agree that it is off-topic) is absolutely, entirely irrelevant to WP:V (but highly relevant to DUE and EP). If some idiot adds a well-sourced summary of Ancient astronauts to Myocardial infarction, you're not going to revert it with an edit summary that says "Rv unverifiable material", because the material is verifiable. Instead, you're going to revert it with an edit summary that says something like, "Rm off-topic, WP:UNDUE, irrelevant information."

We should not be trying to make WP:V do the jobs of these other pages. WhatamIdoing (talk) 14:53, 28 July 2011 (UTC)

Well said. Blueboar (talk) 21:22, 28 July 2011 (UTC)

Indeed. --Nuujinn (talk) 00:01, 29 July 2011 (UTC)

Blueboar, a partial fix is to add to the list of "other policies" a link to WP:EP. 

FYI, I am not objecting, but I question claiming what "Wikipedia strives" to do, haven't we said over and over that WP:V is not the place to define Wikipedia?  And I think it is not preferred as technical writing for the Wikilink to have meaning that is not in the open text, but this particular case seems to be a political compromise.  Unscintillating (talk) 09:00, 28 July 2011 (UTC)

If you mean saying: "...other policies, such as WP:Neutral point of view and WP:Editing policy, also apply"... I would have no problem with that. Blueboar (talk) 12:14, 28 July 2011 (UTC)

Two notes. I think that S Marshall is absolutely right, in the 80-90% of Wikipedia that works, editors routinely exclude sourcable material because it is wrong or false. Second, I wish folks would stop implying that these efforts of ours are to inject new off-topic stuff. IMHO the efforts are all just to try to mitigate the damage being caused by two very off-topic words ("not truth") that are in the policy. The bolded stuff above summarizes it well. North8000 (talk) 14:54, 28 July 2011 (UTC)

I'm pretty close to agreeing with Blueboar's change above, perhaps with the tweak of unlinking "accurately" and putting the link after NPOV. As I think about it, I also agree with what WhatamIdoing said: that "addressing the issue" is likely to undo the meaning of this policy, as it has always existed. I think that if we say that the threshold is verifiability not truth, but we expect you to remove what you think is untrue, then we might as well put this entire guideline up for deletion. But, on the other hand, something else occurs to me. Think of V as being "the threshold": it's the first thing you have to pass, and if it's unverifiable, it cannot be added and it ends right there. But if it passes V, it has passed "the threshold", but hasn't yet fully gotten "into the building". It's like passing the first step, so now there are more steps to pass. That's where "truth" comes into play. Fail verifiability, and the story is over. Pass verifiability, and then the next step is the need to pass truth. This page is about the first step, the threshold. Other policies cover the subsequent steps, truth. --Tryptofish (talk) 21:21, 28 July 2011 (UTC)

As an aside... S Marshal raised the example of reverting out discussion of Sarah Palin's Paul Revere quote from the Paul Revere article... because the editor said Palin was an unreliable source. Having looked at the edit... and the discussions on the talk page... I would completely agree with both removal and the reason... but... I disagree with S Marshal's contention that "the editor removing it has made the value judgment that Sarah Palin was wrong." That may be S Marshal's impression, but it is not backed by any evidence. The editor does not even mention the issue of whether Palin was right or wrong (true/false, accurate/inaccurate, etc.)... he simply says she is an unreliable source. I agree with that ... Palin is a politician not a historian. And in a history related article, the reliable sources are historians... not politicians. What is important to understand is that the unreliability relates to who said it... not what was said. If some eminent historian had said the exact same words that Palin said (in case you don't remember, this relates to her saying that Revere "warned the British"), we would not have excluded it for being "wrong"... and we probably would have to include it as a minority opinion from an eminent and reliable historian. No, the material was not removed because Palin was "wrong", it was (correctly) removed because Palin is not an expert in history. I would argue that even if Palin had been right (and said Revere warned the colonists), we would not have included her statement in the Paul Revere article... and for the same reason... Palin is not a historian... and thus, not a reliable source on Paul Revere.

I raise this because it will help all of you understand why the various editors here disagree so strongly... now back to your regularly scheduled bickering. Blueboar (talk) 21:21, 28 July 2011 (UTC)

And we can try for consensus again!

I'm an eternal optimist (see also: fool), so here is:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Wikipedia strives to present what those reliable sources, taken as a whole, say, as accurately as possible. Consequently, verifiability is necessary but not sufficient for inclusion. Once verifiability has been established, other policies such as neutral point of view and appropriate utilization of sources also apply. ”

--Tryptofish (talk) 21:29, 28 July 2011 (UTC)

"Once" implies a rigid order for evaluating material, which is inaccurate: Many things are rejected as UNDUE without bothering to figure out whether they're technically verifiable.

Also, RS is officially a guideline, not a policy, so "other policies such as" is misleading. WhatamIdoing (talk) 22:29, 28 July 2011 (UTC)

Agree with WhatamI's analysis... but am ok with this in concept. Blueboar (talk) 22:39, 28 July 2011 (UTC)

Would we prefer: (1) "other policies and guidelines", or (2) delete " and appropriate utilization of sources"? Either way works fine for me. And how about "When verifiability can be established,...:? --Tryptofish (talk) 23:05, 28 July 2011 (UTC)

I think 1) works best, and I would prefer "When verifiability is established,...". But the latter's a nit, and I can accept the two other versions without reservation. --Nuujinn (talk) 00:04, 29 July 2011 (UTC)

I think it would be good to say/include the broader term "other considerations". North8000 (talk) 00:07, 29 July 2011 (UTC)

Pending any future howls of complaint comments, any of those would be fine with me! --Tryptofish (talk) 00:44, 29 July 2011 (UTC)

So my idea would be to change "policies" to "considerations" in the last sentence. . Sincerely, North8000 (talk) 01:12, 29 July 2011 (UTC)

I think that would weaken the sentence. Considerations is a bit to vague for my tastes--our actions are informed first by policies, then by guidelines. Considerations might include essays and personal beliefs. --Nuujinn (talk) 01:40, 29 July 2011 (UTC)

What actual problem is this version intended to fix? It seems to satisfy only those who didn't want the policy changed in the first place, so why alter the policy?—S Marshall T/C 09:22, 29 July 2011 (UTC)

Compared to what the page says now, it is intended as an attempt to clarify "not truth"—see also the new thread that started below. But of course, you know that. It's a compromise. For those who consider compromise to be unacceptable, the US Congress is that-a-way. --Tryptofish (talk) 16:55, 29 July 2011 (UTC)

Updated, per the comments so far:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Wikipedia strives to present what those reliable sources, taken as a whole, say, as accurately as possible. Consequently, verifiability is necessary but not sufficient for inclusion. When verifiability is established, other policies and guidelines such as neutral point of view and appropriate utilization of sources also apply. ”

--Tryptofish (talk) 16:59, 29 July 2011 (UTC)

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. The phrase "not truth""Not truth" refers to material that has not been published by a reliable source.  Verifiable material may or may not be accurate—other polices and guidelines address accuracy in the encyclopedia, including Wikipedia:Neutral point of view, WP:RS#Reliability in specific contexts, and WP:Editing policy. ”

  Unscintillating (talk) 02:35, 30 July 2011 (UTC)

• Unscintillating's version seems circular to me. If we define "truth" as that which appears in reliable sources, then the threshold for inclusion is truth. See?

  Tryptofish's seems to add words without addressing any of the problems.—S Marshall T/C 08:36, 30 July 2011 (UTC)

This exchanges of drafts is friendly and constructive, but may not be making progress by not dealing directly with the core issue(s) Maybe S Marshall and I and anyone else in the "we need real change" camp should put our heads together and state the gist of what we are looking for in a compromise....the compromise in the framework of leaving the first sentence as-is and then adding a second sentence or two after it. Then we could state it here and see who is fundamentally for, negotiable and against it. I think that this would help clarify the situation. If there are folks who are not willing to take a step towards the middle ground, I respect that, but then will know that and recognize that they will support any draft that does so. North8000 (talk) 14:03, 30 July 2011 (UTC)

Sounds like a good plan. I would ask that you all include 3-4 links to discussions that got bogged down for a long time where you all feel that editors were arguing that something they believe was not true but verifiable should be included in an article, including that 10,000 word conversation, as that will help use understand whence you all come to this issue. --Nuujinn (talk) 14:21, 30 July 2011 (UTC)

(added later) Nuujinn, for various reasons (mostly because of how much I have said about it), I will give you the ~10,000-word-conversation-to-remove-one-false-word location (how to you link to ~10,000 words over months?) privately but not publicly. North8000 (talk) 16:29, 30 July 2011 (UTC)

That's fine, pop me an email with a link to the page start of the discussion, and I'll try to trace it out from there. I'll try to kept any comments about it general so as to not give away the actual discussion, and will not complain if you or an admin redacts any comments that seem untoward, in deference to your desire for privacy. --Nuujinn (talk) 16:45, 30 July 2011 (UTC)

Cool. Actually desire to not re-ignite a difficult situation where I have hopes of things going better there eventually. North8000 (talk) 16:53, 30 July 2011 (UTC)

Done. I was afraid I might get chided for guessing high, but now I checked and its over 14,000 words. North8000 (talk) 21:51, 30 July 2011 (UTC)

Yes, links to actual debates that centered on this issue will help. Blueboar (talk) 14:31, 30 July 2011 (UTC)

I, too, am happy to consider empirical evidence that would potentially make me reconsider, but pending that evidence, I have to oppose Unscintillating's version as, in effect, deprecating "not truth" to an extent that does not have community consensus. And I want to point out—strongly!—to the big change camp that we have another alternative, which is to leave the page exactly as it is. That would be fine with me, and would marginally reflect the consensus of the community. Don't waste time and effort trying to get something that you will not get. --Tryptofish (talk) 16:54, 30 July 2011 (UTC)

In response to reasonable and substantial mis-readings by both S Marshall and Tryptofish, I have changed the words "Not truth" to "The phrase 'not truth' ".  I hope this clarifies that there is no big change or the defining of "truth" being proposed here.  Unscintillating (talk) 17:14, 30 July 2011 (UTC)

Thanks, but that doesn't resolve my concerns, nor change my position. --Tryptofish (talk) 17:19, 30 July 2011 (UTC)

And your position is that there is a "big change camp"?  Unscintillating (talk) 17:49, 30 July 2011 (UTC)

Verifiable material may or may not be accurate, agree?

S Marshall argues to circular logic regarding truth in a proposal that removes "truth" from this policy within the limitation that "not truth" must remain.  North8000 argues to identifying core issues, when a core issue is that WT:V is (inappropriately IMO) entangled in "truth" arguments.  Do we agree that verifiable material may or may not be accurate?  Unscintillating (talk) 16:51, 30 July 2011 (UTC)

However well intentioned, this question strikes me as unlikely to get us anywhere. Verifiable information can, indeed, be inaccurate, as is the belief that the community is willing to remove or neutralize the phrase "not truth". --Tryptofish (talk) 16:57, 30 July 2011 (UTC)

For someone that agrees with the question, why talk against it?  Are you trying to obstruct consensus?  Obstructing consensus building is considered uncivil.  Unscintillating (talk) 17:14, 30 July 2011 (UTC)

No. --Tryptofish (talk) 17:19, 30 July 2011 (UTC)

"Verifiable" on Wikipedia means "reported in reliable sources". It's a given that normally reliable sources can be wrong, whether through honest error, being hoaxed or malicious lies. Therefore something can be verifiable but inaccurate.—S Marshall T/C 17:21, 30 July 2011 (UTC)

Yes, and I note that in the cases you cite, the reason we know that a reliable source was wrong is that the error was reported in, drum roll please, reliable sources. That's how we verify that an error occurred. Absent verification in reliable sources, there is simply no way to gauge the accuracy of statements in WP. Relying on our version of the Truth is OR. That's why we say verifiability, not truth. --Nuujinn (talk) 18:56, 30 July 2011 (UTC)

Sorry, but I'm not aware that sarcasm is a vector force in the force of reason.  If you look at 1930 Palm Island tragedy, you will see that reliable sources give both the name "Prior" and "Pryor".  No "reliable source" says that "Pryor" is wrong, but the fact that only the early references to the name use that spelling was enough is assign less WP:DUE weight to the spelling "Pryor".  A second source, Prior's nephew, has also testified in edit comments that "Pryor" is wrong.  So even though we can't repeat this assertion in the article, we have a fair degree of confidence in how to weight the choice between the spellings.  Unscintillating (talk) 19:31, 30 July 2011 (UTC)

Sorry, but I wasn't being sarcastic. My point is no matter how we slice it, we have to verify "that which is True" and "that which is not True" against reliable sources. Regarding Prior/Pryor, I don't see what your point is. We normally give more weight to later accounts in historical matters, so what's the problem? Changing the wording of V would make no difference in that case. --Nuujinn (talk) 20:37, 30 July 2011 (UTC)

Nuujinn, I think you're inadvertently making my point for me. Did you understand that I'd be perfectly happy to define "truth" in terms of reliable sources? But my answer earlier was in terms of what the policy says at the moment. At the moment the policy draws a distinction between what's sourced and what's true. It says "verifiability, not truth". Therefore, if the policy is written correctly, "verifiable" cannot possibly mean the same thing as "true", can it?—S Marshall T/C 21:06, 30 July 2011 (UTC)

We are going in circles again. Verifiability is not the same as Truth. Truth is based on belief... Verifiability is based on reliable sources. WP:V makes it clear that we do not even consider inclusion of material based on an editor's belief ... We do consider inclusion of material that is based on reliable sources (But note... the fact that we consider inclusion does not mean we guarantee inclusion... because there are other policies and guidelines that might tell us to reject it). Blueboar (talk) 22:14, 30 July 2011 (UTC)

You may believe that truth is based on belief, but I don't necessarily believe that. :)—S Marshall T/C 22:51, 30 July 2011 (UTC)

So somehow "drum roll please" was not sarcastic, and somehow capitalizing the word "truth" was not sarcastic.  What I'd really rather know though is: is it your position that both the spelling "Pryor" and the spelling "Prior" are without error, because no reliable source has identified one of them as erroneous?  Unscintillating (talk) 22:16, 30 July 2011 (UTC)

I have a different sense of humor than you. Drum roll please, because no matter how we slice it, discussions of Truth here wind up as discussions of verifiability, and we've been through that particular loop, I don't know, a couple of dozen times the last few months. "T"ruth, because philosophically I fall somewhere between the pragmatists (Pierce in particular) and the european phenomenologists (Husserl, Ingarden, Merleau-Ponty) and Wittgenstein. I believe in little "t" truth, where truth is something that helps solve a problem in a given context. Big "T" truth is the way most people use it, as in "A is true in all cases all the time". As I walk, the earth is usually flat. For an engineer building a road through a mountain, it's not. From space, it appears as a disc. For astronomy, it's the spheroid. So, for me, the sky is WP:BLUE and the sky is WP:NOTBLUE, depending on the circumstances and context. In regard to Prior/Pryor, we have conflicting sources, so we document the difference, and to me, it doesn't matter which it the truth. So that article looks fine to me. And S Marshall, verifiable and true are not synonyms, as Blueboar aptly put it. --Nuujinn (talk) 22:57, 30 July 2011 (UTC)

I began this section by proposing that this overall discussion was inappropriately entangled in "truth" arguments.  I asked what I thought was an easy question, which question is the title of this subsection.  While consensus was not achieved, sufficient consensus was found to draw the conclusion that Wikipedia does not strive for accuracy, at least not as a part of WP:V.  Regarding the question posed, most editors agreed that verifiable material may or may not be accurate, but Nuujinn required a reliable source to tell him/her that one of the spellings ("Pryor", "Prior") is inaccurate.  So either by Nuujinn's ("accuracy is determined by sourcing that may or may not exist") or by most editor's ("some verifiable things are inaccurate") we have a consensus that Wikipedia does not strive, at least as a function of verifiability, for accuracy.  Unscintillating (talk) 22:02, 31 July 2011 (UTC)

Returning to the previous discussion

Returning to the previous discussion, there are two possibilities that look reasonably viable to me:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. Wikipedia strives to present what those reliable sources, taken as a whole, say, as accurately as possible. Consequently, verifiability is necessary but not sufficient for inclusion. When verifiability is established, other policies and guidelines such as neutral point of view and appropriate utilization of sources also apply. ”

or:

“ The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true. ”

I like them both. Either one would be fine with me. --Tryptofish (talk) 20:27, 31 July 2011 (UTC)