Wikipedia talk:Mediation Cabal/Cases/27 February 2012/Wikipedia:Verifiability/Group 4/Archive 1

Archive 1

Comments on individual drafts

Comments on draft 0

Things that are good about draft 0

• I think its good points are apparent except for possibly one: the meaning of "verifiable" is simply the dictionary definition.
• The first sentence indicates the importance of the policy and may increase the reader's attention.
• The second sentence is very quotable for use in cases where an editor tries to put in unpublished original research, true or not.
• The Wikipedia jargon is very minimal, being mainly in the titles of other policies, which is unavoidable.
• The policy is simply stated. The sentences are uncomplicated.
• Simple organization that is separated in appropriate topic paragraphs.

para 1 — must be verifiable

para 2 — inline citations

para 3 — removal of noncompliant material

para 4 — removal of V-compliant material by other policies and guidelines

--Bob K31416 (talk) 20:55, 27 March 2012 (UTC)

• Comment  Tight writing, uses words balanced in their context.  Unscintillating (talk) 03:02, 28 March 2012 (UTC)

Things that could be improved with draft 0

• How did we get stuck with writing wording for WP:BURDEN in this RfC?  The sentence "You may remove any material lacking an inline citation to a reliable source." lacks the context that exists for the same sentence at WP:BURDEN, and it is not the case that newbies or anyone else can in general "remove any material lacking an inline citation".  I'd suggest that even the BLP part doesn't need to be mentioned in the lede.  Unscintillating (talk) 01:16, 28 March 2012 (UTC)

The actual two sentences that I think you are referring to are in paragraph 3,

"Any material that requires an inline citation but does not have a suitable one may be removed. Unsourced contentious material about living people must be removed immediately."

The context is set by the previous two paragraphs in the Draft.

The above two sentences are nearly the same as the corresponding two sentences in the lead of the present version of WP:V.

"Any material that requires a citation but does not have one may be removed. Unsourced contentious material about living people must be removed immediately."

The main changes for the present Draft0 are adding the words "inline" and "suitable". --Bob K31416 (talk) 02:19, 28 March 2012 (UTC)

• The main thing that I think is missing from Draft 0 is an explanation of why we have a verifiability policy. The chief reason is that many of our editors are totally anonymous, and nearly all the rest use unverified names and unverified credentials. So we use verifiability to make our articles approach the truth more closely than if we let anonymous editors write whatever they want with no restraints. How can we state this succinctly (and powerfully)? Jc3s5h (talk) 13:25, 28 March 2012 (UTC)

I didn't think of your suggestion before and it seems original. My initial impression was that it wouldn't work well in the lead. Then for guidance, I checked the leads of NOR and NPOV to see if they had any discussion of similar ideas and they didn't. --Bob K31416 (talk) 18:39, 28 March 2012 (UTC)

• I disagree with "Verifiability is the foremost requirement in Wikipedia." It is important, but no more important than obedience to copyright law or neutral point of view. I would write something like Verifiability is a mandatory requirement in Wikipedia. Jc3s5h (talk) 13:31, 28 March 2012 (UTC)

I agree that all of the policies are important requirements. Any one of them alone can prevent material from being added to Wikipedia. But I think Verifiability is the most fundamental, of most use, and is the soul of Wikipedia because it does the most to promote accuracy and credibility. --Bob K31416 (talk) 18:39, 28 March 2012 (UTC)

Other arguments could be made. For example, one might claim NPOV the most important because verifiable reliable sources could be presented in a one-sided way to create an unfair article. Once could violate the copyright policy to create an article that actually violates the law. Unverifiable articles, viewed by an outsider unfamiliar with Wikipedia policies, are only potentially problematic; sometimes unverifiable articles are accurate. Articles that violate NPOV or copyright would always be considered bad articles by a well-informed unbiased reader, even if the reader was not familiar with WP policies. Jc3s5h (talk) 19:07, 28 March 2012 (UTC)

I think you make intelligent points, but I don't think they are comprehensive enough to be convincing for me.

I think that in most articles in Wikipedia, NPOV is not an issue. You may disagree. I suppose we could each check that for ourselves by using the random article generator at the upper left of the page. The copyright issue can come up on any article, but in my experience it's not that much. The issue of Verifiability, on the other hand, applies to all articles and seems to come up the most.

In the long-standing version of WP:V the first sentence started with, "The threshold for inclusion in Wikipedia is verifiability, ...". The predominant interpretation of this is that first the material has to pass the verifiability test and then it has to pass the other tests. Agree with it or not, it seems that a large majority of editors who have discussed it at WT:V seem to believe that. This is consistent with Verifiability being the foremost requirement. --Bob K31416 (talk) 23:51, 28 March 2012 (UTC)

I have indeed noticed that some editors seem to have a mindset that the first thing, chronologically, in the article creation or review process, is verifiability. I disagree with that. Editors are free, while creating an article off-line, to think about, or revise for, any of the policies in any sequence they see fit, so long as the finished article meets the policies (or at least isn't so bad that it must be deleted immediately). Similarly, administrators could delete an article, or any editor could delete an addition, for any policy violation, and are free to review an article or addition for any kind of violation in any sequence they wish. I view a new lede draft as an opportunity to eradicate this spurious sequence notion.

If verifiability really does constitute the greatest quantity of violations, perhaps we could word it to indicate specifically that it is the greatest quantity, rather than most important or first in sequence. Jc3s5h (talk) 15:23, 29 March 2012 (UTC)

Although we have some differences, do we both have the same impression that verifiability constitutes the greatest quantity of violations? --Bob K31416 (talk) 16:03, 29 March 2012 (UTC)

Yes, it is my impression that verifiability violations are the greatest quantity of violations, in two senses. First, I think it is the greatest when counting passages in violation. Second, I think it is the greatest when counting the number of articles that ought to be deleted if there is no improvement. But I'm willing to change my mind if someone wants to count some articles. Jc3s5h (talk) 16:30, 29 March 2012 (UTC)

Thanks. I'll pass on counting articles for now. : ) Here's some other thoughts that I was adding to my previous message when I got an ec and I stopped to edit them some more:

Verifiability does seem to come first in various ways.

1. I think that most of the time, editors are aware of the Verifiability policy and see info in reliable sources and then put it into Wikipedia. Then, for example, NPOV is considered when the material can be seen in the context of the Wikipedia article.

2. Copyright violations are usually found by first challenging the material that violates Verifiability and doesn't have a citation. When the citation is put in, then they can see that it is a copyright violation. Or if there already is a citation because of the Verifiability requirement, then it can be seen that there is a copyright violation.

3. Similarly, OR is usually concluded when a citation is requested, according to Verifiability, for dubious material and it's not forthcoming, or the citation does not explicitly support the material.

4. Evidence that an article is not NPOV can be found when the citations required by Verifiability indicate mostly sources of a particular political persuasion.

It seems that compliance with Verifiability leads to discovery of violations of other policies or compliance with other policies in many cases, whereas compliance with other policies doesn't lead to compliance with verifiability. --Bob K31416 (talk) 16:52, 29 March 2012 (UTC)

Your list seems to be largely from the point of view of someone who is reviewing articles, whether formally as part of a good article or featured article review, or for the editors own satisfaction. But one good editor, knowledgeable in the field, might write a draft based on his own knowledge, which would inherently comply with copyright, and the editor would probably try to have a neutral point of view. In this case, the last step would be to find citations to make the article verifiable.

In another instance, an editor might decide that the information in a certain expensive copyrighted publication ought to be made available free. So the editor might start with the copyrighted pub, pare it down, summarize it, restructure it, bring in other points of view, and finally find sources. So the sequence would be make it free of copyright violations, then NPOV, and finally V.

When considering problem material, we should not only consider the articles that made it, but the ones that were submitted by IP editors that were never accepted, and the ones that were deleted so quickly most of us never noticed them. I think in those we might see a great number of notability, advertising, copyright, and NPOV violations. Jc3s5h (talk) 17:13, 29 March 2012 (UTC)

I thought your arguments were using very atypical situations with regard to all the edits that are made on Wikipedia each day. I've been trying to find reasons for agreeing with your position but all that has happened is that I found more reasons for supporting the use of "foremost". I guess we need other editors to jump in here.

Anyhow, you've had some thoughts about changing the first sentence of this draft. Could you give here your suggested version of the first sentence? --Bob K31416 (talk) 17:29, 29 March 2012 (UTC)

Here's a possibility for a sentence to replace the present first sentence of the draft that may satisfy both your previous suggestion of having "why verifiability" and your present comments about "foremost.

"Verifiability is essential for the credibiity of Wikipedia."

--Bob K31416 (talk) 14:57, 30 March 2012 (UTC)

I agree with those who think the first sentence is a can of worms. How can WP:V be foremost, making it more important than other content policies? They interlink and make no sense without each other.--Andrew Lancaster (talk) 14:21, 31 March 2012 (UTC)

Andrew, The previous opening in WP:V was, "The threshold for inclusion in Wikipedia is Verifiability...". That seemed to put Verifiability as foremost. Did you object to that special status for Verifiability back then and if so what did you say? --Bob K31416 (talk) 02:27, 1 April 2012 (UTC)

Comments on draft 1

Draft 1 rationale

The inspiration for draft 1 comes from the tight writing and the text of Draft 0.  I've excised as much material as I could from the lede, while trying to attend to good factoring.  For example, I removed the reference to WP:Published.  It is neither a policy, nor a guideline, nor an essay.  If that should go back in, it belongs lower on the policy page.  The entire clause "published, reliable sources" has been made a single Wikilink to a section lower on the policy page.  Likewise, the necessity for citations, and the WP:BURDEN of sourcing, are each linked to a target lower on the page, but this time using open citations.  I think we need a policy-page section that lists the core content policies with their common names, while there is no need for this in the lede.  WP:NOT is relevant in a mention of key content policies.  The most vital claim for the need for VnT reappears in the subsection "Perceived truth and personal experience".  Unscintillating (talk) 01:12, 30 March 2012 (UTC)

Things that are good about draft 1

• Draft 1 includes an explanation of why the policy exists—to insure Wikipedia is not the first publication of an idea.
• Draft 1 does not imply any temporal or importance ordering of the policies.

Things that could be improved with draft 1

• Copyediting: The standard typography for an encyclopedia article or heading within an article is to enclose it in double quotes: "Gregorian calendar". I think it's reasonable to extend this convention to pages and headings in project space, like "What Wikipedia is not." Jc3s5h (talk) 12:18, 30 March 2012 (UTC)
• Is Draft 1 the lead or the lead and two additional sections (Core content policies and Perceived truth and personal experience)? If the former, you may want to change the formatting. --Bob K31416 (talk) 14:13, 30 March 2012 (UTC)

They are supposed to be sections that use "==", but because of the formatting on the project page I had to use three "===".  As stated in the rationale, "I think we need a policy-page section that lists the core content policies with their common names, while there is no need for this in the lede."  The section about "perceived truth" IMO isn't something needed in the lede.  A couple more things that could be removed, WP:BURDEN doesn't need to be mentioned in the lede, and WP:Copyright isn't strictly a content policy.  Is there a clear relationship between WP:V and WP:Copyright?  Unscintillating (talk) 23:04, 30 March 2012 (UTC)

Thanks. May I suggest that you put something like the following sentence after the lead and before the sections to clarify for editors who will be looking at the work.

"[The following two sections are not part of the lead but are proposed to be included in the policy after the lead.]"

--Bob K31416 (talk) 02:39, 31 March 2012 (UTC)

• It doesn't look like the struck out parts of the following sentence are needed.

Verifiability means that the ideas don't originate at Wikipedia, but instead are traceable to published, reliable sources.

Also, you might consider changing "ideas" to "information" etc.

--Bob K31416 (talk) 02:48, 31 March 2012 (UTC)

I'll prepare another draft with the excisions.  As for "ideas", I'd have to change that to "material".  The problem with "information" is that there are a lot of things that are "information" in an article that are not, and should not be, verifiable.  Like adding the footnote that gives the alternate spelling "Pryor" is information that Wikipedia thinks that the "Pryor" spelling is less important than the "Prior" spelling.  I think that WP:DUE decisions in general come from editorial judgments, and are not traceable to a reliable source.  "Information" also includes word count per paragraph, style of writing, and reading level.  Unscintillating (talk) 03:15, 31 March 2012 (UTC)

Comments on draft 2

Draft 2 rationale

Draft 2 is a revision of Draft 1.  It incorporates previous suggestions, mostly removing removable text.  After asking myself why I added WP:NOT in Draft 1, I reversed direction, and re-characterized the concept for the section as "Three related core content policies".  Unscintillating (talk) 04:09, 31 March 2012 (UTC)

Things that could be improved with draft 2

• The use of the verb relating in the third sentence is a bit trendy but vague. I am sure that can be better worded?
• The last section about perceived truth could possibly make it specific that this concerns inclusion. Perceived truth can be relevant when discussing what not to include.--Andrew Lancaster (talk) 14:25, 31 March 2012 (UTC)
• I think the above also applies to draft 3, which no one has commented on.--Andrew Lancaster (talk) 14:48, 31 March 2012 (UTC)

the word "relate"

The concern is that the word "relate" is trendy, vague, and can be better worded.  It was not written to be trendy or vague, but as the technical operational definition for the previous word "traceable".  Perhaps the word "associate" would be more clearly a technical information-technology term.  Unscintillating (talk) 19:48, 31 March 2012 (UTC)

does the statement about perceived truth discourage the use of perceived truth in other policy contexts

The concern is that asserting that perceived truth cannot be used as a substitute for the verifiability requirement, could be mis-associated to mean that perceived truth cannot be used in other policy contexts.  In another draft, I characterized verifiability as a "constraint".  Another approach would be to add another sentence, "Verifiable material may or may not be accurate, [[WP:DUE|significant]], compliant with the '[[Wikipedia:What Wikipedia is not]]' policy ('[[WP:NOT]]'), or compliant with the '[[WP:Copyright violation]]' policy ('[[WP:COPYVIO]]')."  Unscintillating (talk) 19:48, 31 March 2012 (UTC)

Comments on draft 3

Withdrawn. --Bob K31416 (talk) 12:46, 1 April 2012 (UTC)

Comments on draft 4

Draft 4 rationale

I have pasted this in based on my own thoughts, but I started out by using North8000's draft from step 2, which was my vote in that round. I see it as an attempt to get back to basics and really write from the ground up while taking note of what most people seem to me to find most important.--Andrew Lancaster (talk) 14:51, 31 March 2012 (UTC)

Things that could be improved with draft 4

• Re "It is not enough that the information is true." — This still contains the notion that truth isn't important. I also note that the link goes to WP:Verifiability, not truth. Using "true" is a can of worms. It seems that the motivation for this part is compromise with VnT advocates, which means that this draft may be better placed in another group. --Bob K31416 (talk) 15:34, 31 March 2012 (UTC)

First, not it does not say that truth is unimportant. I do not take that position, and I did not write that way. Second, trying to remove all mention of the concept truth might well be impossible, not just because the community will not accept it, but also because the good part of the VnT idea (I think) is to tell people what to do when they think something is true. How much emotion gets burned up by editors when they think something is truth and they can not understand why people are reverting their edits? Such incidents need the help of a good policy page.--Andrew Lancaster (talk) 20:20, 31 March 2012 (UTC)

I think this sentence is a backdoor way of putting in policy that material must be true.  This sentence has already led perhaps five people to claim that, in various ways, "Dewey Defeats Truman" is true, because their basic alternative is to say that the article "Dewey Defeats Truman" must be removed from the encyclopedia.  The approach that the material is not reliable in the context also doesn't work, because the three-word clause then fails WP:V, and cannot be included in the encyclopedia.  I think it is not a difficult problem once we accept that "Dewey Defeats Truman" is verifiable, untrue, and more than WP:DUE insignificant.  Unscintillating (talk) 20:50, 31 March 2012 (UTC)

• "other policies, guidelines and considerations apply." — Which policies, guidelines, and considerations apply?  Unscintillating (talk) 20:50, 31 March 2012 (UTC)

• Mention of inline citations is delayed. --Bob K31416 (talk) 00:31, 1 April 2012 (UTC)

• In the second paragraph, the second sentence is more important than the first and should start the paragraph. --Bob K31416 (talk) 00:31, 1 April 2012 (UTC)

Comments on draft 5

Draft 5 rationale

The main change from Draft 2 is to add "verifiable material may or may not be accurate" along with related factoring.  A "constraint" is a "restricting condition", which should help give perspective to what verifiability is.  Swapped out "relate" with "link".  I went back to using italics for wikilinked policy titles, and also introduced italics for technical terms defined with Wikilinks. It is a minor point if the MOS people prefer quote marks, but to me the italics communicate better than the quote marks. Historic basketball game on TV right now in the NCAA Final 4, Louisville vs. Kentucky!  Unscintillating (talk) 22:30, 31 March 2012 (UTC)

Things that could be improved with draft 5

Archive 1