User talk:Crum375/att

This is the talk page for discussing the draft proposal of a new WP:ATT policy. Before commenting here, please read the rationale page.

Links

Source pages

Initial draft

I started an initial draft for a new WP:Attribution policy, using as input the current WP:V, WP:NOR, and WP:ATT, as well as many of the recent comments on WT:NOR and WT:V. Comments welcome. Crum375 (talk) 22:03, 23 May 2010 (UTC)[reply]

Note that the flow and organization are somewhat clunky, since much of the material was imported verbatim from WP:OR and WP:V. I felt that, wherever possible, it is important to preserve existing content first, and remove or condense it only with community input. Crum375 (talk) 22:10, 23 May 2010 (UTC)[reply]

Re Primary Sources

I would add something to this paragraph to the effect: Primary Sources that are widely cited in secondary or other primary sources are generally suitable and reliable for citing specific facts, ideas or quotations in an article. I think this has much greater validity the older a primary source is. Primary sources written in the 17th-early 20th century that have been widely cited in other works, should be condsidered reliable.--Mike Cline (talk) 22:32, 23 May 2010 (UTC)[reply]

To avoid controversy, I copied the current WP:PSTS section essentially verbatim. If substantive changes are proposed to PSTS, I think they should either be made on WT:NOR, which is the current talk page for that policy, or wait until this policy is adopted, which may take a while. Crum375 (talk) 22:43, 23 May 2010 (UTC)[reply]

You might also want to take a look at the Proposed guideline at Wikipedia:Primary, secondary and tertiary sources, which could be relevant to your project. --SteveMcCluskey (talk) 19:32, 24 May 2010 (UTC)[reply]

As I recall, that proposal failed to gain consensus, and my goal here is to take the existing official policy sections and not introduce substantive changes to current policy until and unless this proposal becomes policy. At that point, any section, including PSTS, can be modified by consensus. Crum375 (talk) 19:39, 24 May 2010 (UTC)[reply]

Attribute, attributable, attribution. Right words?

I googled using the keywords: 1) attributable definition 2) attribution definition 3) attribute definition. Are you sure that these are the right words to use? --Bob K31416 (talk) 01:46, 24 May 2010 (UTC)[reply]

What do you see as a problem? Crum375 (talk) 01:56, 24 May 2010 (UTC)[reply]

After reading the definitions by googling the above, you are satisfied with those words? Hmmm.

Anyhow. You might compare the definitions of the verbs attribute and cite using the following links:

attribute (Also note, "used as a verb is rare".)

cite (Also note, "used as a verb is common".)

Maybe WP:CITE would be better than WP:ATTRIBUTE? --Bob K31416 (talk) 02:21, 24 May 2010 (UTC)[reply]

To the best of my knowledge, in the publishing business (which WP is in) the verb "attribute" is well defined and commonly used. We have been using it on WP in the same publishing sense for years. To "cite" is a synonym in a sense, but I believe that "attribute" is more appropriate, esp. when we include the forms "attributable", "attributability", "unattributable", etc. Also, on WP the terms "cite" and "citation" refer to the physical format of the source description (e.g. citation template, Harvard citation), while "attribution" refers more to the concept behind it. But the bottom line is that "attribute" in its various forms is well defined and common both in the publishing business and on WP. Crum375 (talk) 02:33, 24 May 2010 (UTC)[reply]

Very, very few editors that come to Wikipedia are in the publishing business. When they come to policy pages to get clarification of the reason that one of their edits was reverted, shouldn't the policy pages be written for them, using words that they are the most familiar with? --Bob K31416 (talk) 02:40, 24 May 2010 (UTC)[reply]

Perhaps very few come here from the publishing business, but once they are here, they are in the business of publishing, so that terminology is the most relevant. Also regarding "familiar" terms, "to cite" is often used as a synonym for "to quote". So a citation (in publishing) could mean a quotation of a source, which would be confusing. Also, the more common meanings of "citation" are a commendation, or a summons, both of which could add confusion if used in lieu of "to attribute" and "attribution". The noun "attribution" is clear and well-defined, it means supplying a source which makes (or directly supports) the statement that we publish. "Citation" is more amorphous. Crum375 (talk) 02:50, 24 May 2010 (UTC)[reply]

It might be helpful if you cited here a source for the definition of attribute that you want to use and give the excerpt here, and perhaps you could comment as to how rare the definition is. Note that the link I provided said that the use of "attribute" as a verb was rare, and to use it in the sense that you want to use it, would thus be even rarer. --Bob K31416 (talk) 03:02, 24 May 2010 (UTC)[reply]

Like all English words, the meaning depends on context, so in the publishing context it would be meaning #3 for the infinitive: "at·trib·ute (transitive): 3. to consider as made by the one indicated, esp. with strong evidence but in the absence of conclusive proof: to attribute a painting to an artist."

To the best of my knowledge, in the publishing context, not only is it not rare, but it is the most appropriate (and possibly the only) word which means to supply a source for a published statement. Crum375 (talk) 03:14, 24 May 2010 (UTC)[reply]

Let me emphasize another linguistic point: "To cite a source" means to quote (or provide) a source. On the other hand, "to attribute a statement to a source" has two objects: the statement and the source, and it makes the connection between them. In other words, it means that the source made that statement (although it might be paraphrased). As far as I know, "to cite" is not the correct term for this usage, and "to attribute" is the only available choice. Crum375 (talk) 03:26, 24 May 2010 (UTC)[reply]

The definition from the source that you cited, has the meaning of giving someone credit for something, i.e. to judge that someone deserves credit especially "in the absence of conclusive proof". Are you sure you want to use that word, attribute? Note how much more appropriate is definition #2 for cite from the same source that you used.

2. to mention in support, proof, or confirmation

But hey, it's your project and you need to use what you feel you can defend. Good luck. --Bob K31416 (talk) 03:41, 24 May 2010 (UTC)[reply]

No, I think there is only one English word which means in a publishing context "to indicate the source for a statement", and it is "to attribute". "To cite" means to quote (or indicate) a source (in a publishing context), but it does not make the connection between the statement being published and the source which made it. So cite is simply not the correct choice, and on WP itself it's used to quote or supply a source (independent of the statement made). In summary, "attribute" is the only correct linguistic choice, "cite" is incorrect for this use and conflicting with other use on WP, and WP editors (and policies) have been using "attribute" and "attributability" correctly for years. Crum375 (talk) 03:50, 24 May 2010 (UTC)[reply]

The word attribution does not appear in a great many publishing glossaries online. That said I've always known what attribution meant. My suggestion is to leave the terms attribution, attributable and unattributable in the guideline because they are nouns and adjectives/adverbs with an uneqivocal meaning. On the other hand, I would substitute the word cite for every instance of the verb attribute. For example the sentence: The only way to demonstrate that material is not unattributable is to attribute it using an inline citation to reliable.... would read better if it read: The only way to demonstrate that material is not unattributable is to cite it using an inline citation to reliable..... There aren't many instances of the verb attribute in the guideline, but changing it to the verb cite will make for a better guideline.--Mike Cline (talk) 00:42, 25 May 2010 (UTC)[reply]

As I noted above, using the verb "to cite" can introduce confusion. In its publishing context, "to cite X" normally means to quote source (or reference) X, or provide a link to it, independent of any specific content statement that needs sourcing. On the other hand, the verb "to attribute" is a "connecting verb", in the sense that when we say "to attribute X to Y", it means to provide content statement X with a source Y. This is exactly what we need on Wikipedia: every content statement X needs a reliable source Y, whether we actually provide it ("attribute X to Y") or not ("X is attributable", but Y is not specified). Because of this difference, when we use "cite" (or "citation") on WP we normally, but not always, focus on the source (or reference), and the physical style and formatting of that reference. On the other hand, "to attribute" is unambiguous, and it always refers to supplying a source in support of a specific content statement, per the attributability requirement. Crum375 (talk) 01:27, 25 May 2010 (UTC)[reply]

Mike Cline, Re "The word attribution does not appear in a great many publishing glossaries online. - Could you give links to any of those publishing glossaries where attribution appears? Thanks. --Bob K31416 (talk) 15:04, 3 June 2010 (UTC)[reply]

Bob, is this helpful? Or this? And here is what Wikipedia says about it. And the Wikia-Journalism glossary (search for "Attribution"). Crum375 (talk) 15:18, 3 June 2010 (UTC)[reply]

Bob, if you use the terms Publishing Glossary in Bing [1], none of the glossaries on the first page contain the word Attribution. It is not a publishing term. However, Crum's contention that it is a publishing term is being confused with Journalism. Attribution is a common term in the Journalism business, but not the publishing business.--Mike Cline (talk) 16:54, 3 June 2010 (UTC)[reply]

"Attribution" is the only term I am aware of which is used professionally to identify the source of a statement being cited, as the sources I linked to above show.[2][3] On Wikipedia itself, both in article space as well in policies, "attribution" is used as "source identification". Since the goal of this combined policy is to be as faithful to the original ones it derives from, it makes sense to minimize unneeded changes and use the same word. Crum375 (talk) 17:54, 3 June 2010 (UTC)[reply]

Crum - Attribution is the right word, its just not a publishing word, its a journalism word. I was just responding to Bob's question, not questioning the use of the word in the draft guideline. That said, I still think the guideline would read better if you replaced the verb attribute with the word cite in the couple of places it occurs. Doesn't change the meaning or intent one iota.--Mike Cline (talk) 18:12, 3 June 2010 (UTC)[reply]

Mike, can you point out a specific sentence in the draft where in your view "cite" would be better than "attribute"? Crum375 (talk) 18:20, 3 June 2010 (UTC)[reply]

Crum, look above, last comment before this outdent--Mike Cline (talk) 18:53, 3 June 2010 (UTC)[reply]

I've already replied to that one. "To cite" in our context, generally means "to quote (or supply) a source", independent of any content material. On the other hand, "to attribute" means to connect a specific content statement to the source which made it (or supports it), which is what we need on Wikipedia. Alternatively, looking at the noun, "an attribution" is the act of identifying a source for a given statement. In other words, it connects a specific statement to a specific source. On the other hand, a "citation" (ignoring the fact that it can also mean "commendation" or "summons"), refers to the source itself, independent of the content statement. So again, the point is not to supply any old source, but to link a specific statement to a specific source. The verb "to cite" doesn't do that, while "to attribute" does. Crum375 (talk) 19:30, 3 June 2010 (UTC)[reply]

Here's just a general comment that you can take or leave as you wish. I think you have gotten so focused on using words that begin with attribut... that you are unable to put yourself in the position of someone reading what you have done. Your considerable use of words beginning with attribut... does not read well, and the words themselves are rarely used in the english language, and even less in the way that you are using them. And using "attribution" as the title of the proposed policy is the most glaring error of all IMO. The title should be "source" or "cite", and editors should be advised to "cite reliable sources", and it would be better if the words beginning with attribut.. weren't used, again IMO. Regards, --Bob K31416 (talk) 14:19, 4 June 2010 (UTC)[reply]

My goal is to use the existing Wikipedia terminology and minimize changes, except where absolutely necessary. The "attribution" terminology is widely used on Wikipedia talk pages and content policies, as well as by the outside world per my above links. The words you suggest, "source" and "cite" are not connective terms, and don't conceptually link the source to a content statement like the verb "to attribute", so they don't fulfill the same function. Both words are used in the draft and on Wikipedia where the focus is on the source itself and its format, as in Wikipedia:Citing sources. Crum375 (talk) 15:45, 4 June 2010 (UTC)[reply]

I agree that "attribute" is definitionally distinct from "cite," and often the correct choice among the two. Perhaps never the sole word available. The distinction between "journalism" and publishing, is inadequately meaningful in this context.

Calamitybrook (talk) 01:19, 11 June 2010 (UTC)[reply]

No "No original research"?

Offhand I think it would be necessary to include a section on "No original research" in order for this to have any chance at passing muster in front of the community. Perhaps somewhat similarly to what's already at WP:ATT? Given its rich history as one of the three core content policies, I don't think it can realistically be made to just vanish into thin air. ... Kenosis (talk) 00:06, 28 May 2010 (UTC)[reply]

It's not vanished at all. It is simply renamed to "no unattributable material", or "no unattributable content". The WP:NOR shortcut remains the same, and the definitions are the same, with much of the text copied verbatim. The reason for the renaming of NOR to NUC (NUM is taken a the moment) is that the term "original research" creates vast confusion on WP, and is in many ways a misnomer, because by NOR we don't really mean "original", nor do we really mean "research". So the renaming (with no change in the underlying concept) is a key piece of this proposal. If you can find some part of WP:NOR that is missing here, it would be very helpful. Crum375 (talk) 00:32, 28 May 2010 (UTC)[reply]

Please remember that "original research" has a special meaning in academic work that is independent of the issue of reliable sourcing. I commented about this at WT:NOR recently. A primary point of "no original research" is to reject the kind of well-sourced original research that is often expected of graduate students and of many undergraduates as well. Sometimes this is equivalent to what we've come to call "original synthesis" (though it's still often called "original research" rather than "original synthesis" in academia). So there is definitely a point to the choice of the words "no original research". Countless times it's been necessary to explain to editors with an academic background that what they learned to do in an academic paper is not what we do here, and the words "no original research" help make the point to those editors that are thusly oriented. ... Kenosis (talk) 02:43, 28 May 2010 (UTC)[reply]

One key rationale for this proposed draft was the apparent feeling by many editors that the term "original research" causes confusion. The point of this draft policy is not to change the underlying concept, which remains the same, but only its name. For those people who prefer the old name, a shortcut will remain which points to essentially the same thing. Crum375 (talk) 02:55, 28 May 2010 (UTC)[reply]

WT:NOR RfC on changing the definition of NOR

(Please note that this section was created by Crum375 and the first message was a response to Crum375's message of 00:32, 28 May 2010 (UTC) in the above section No original research"? and was originally in that section, along with the 3 messages following it.)

Re "the term 'original research' creates vast confusion on WP, and is in many ways a misnomer, because by NOR we don't really mean 'original' " - But over at WT:NOR you opposed the effort to simply fix that at the RFC here. So if you want to fix that with your proposed policy WP:ATT, why did you oppose fixing it at WP:NOR? --Bob K31416 (talk) 01:28, 28 May 2010 (UTC)[reply]

The confusion that we currently have with NOR is that it has a misleading name, not that the underlying concept is flawed. In fact, the underlying concepts of both V and NOR are just fine, and need no modification. The RfC in question proposed to change the underlying concepts, in a way that, in my opinion, would add even more confusion and would not solve anything that I can see. This proposed draft policy does not intend to change any underlying concept, and instead focuses on presenting the existing concepts, which have served us well, in a way that is better organized and with names that are less confusing, hopefully easier to understand and use. Obviously only the community as a whole can decide whether this combined version is better than the existing pair of V and NOR. Crum375 (talk) 02:00, 28 May 2010 (UTC)[reply]

Re "The RfC in question proposed to change the underlying concepts" - That's false. The underlying concept of excluding material from unreliable sources was still present in WP:VER. The only thing the RFC would have done was fixed the definition of Original Research. --Bob K31416 (talk) 02:15, 28 May 2010 (UTC)[reply]

Changing the definition of a core policy which has been constant for years, and is linked to in countless discussions around the site, will introduce a major change in a core concept and render all those threads meaningless. I don't think that's a step forward, when the only issue is a bad name. But in any case, the RfC and specific NOR discussion belong on the WT:NOR page. Crum375 (talk) 02:29, 28 May 2010 (UTC)[reply]

Discussion continued at WT:NOR[4] per request. --Bob K31416 (talk) 04:36, 28 May 2010 (UTC)[reply]

Truth

I'm curious why you removed the statement: verifiability, not truth. It seems most debates are over editors concepts of what they believe is truth and because their truth, which is essential to their POV, is so dear to them (sometimes literally religious) they will fight to push non-truth from an article. The truths they hold to be self-evident cuts into debates over WP:RS reliable sources. When I read the statement verifiability, not truth a light went off over my head and my understanding of WP:NPOV improved. Verifiability, not truth became an editing motto that allowed me to more easily accept consensus and work to include reliable and verifiable POV's that I might personally find distasteful. On the other hand, if the general public get's told by some demagogue that Wikipedia's mission isn't truth but to politically correctly waffle on it's content then trust in our encyclopedia will be undermined and the recruiting of new editors damaged. So verifiability, not truth becomes a liability. Should truth be used at all in this edition? Alatari (talk) 06:20, 28 May 2010 (UTC)[reply]

Good question. I used to feel about this issue exactly as you do, and for me that point about "not truth" hit home. But having been around for over four years now, I realize that it causes more trouble than it's worth. It's essentially a red flag for many editors, who come charging in to policy talk pages and initiate long talk page threads about why "not truth" is wrong, and why we should get rid of it. So I thought about it and realized that by focusing on the simpler language, "everything must be attributable to reliable sources", we avoid the main pitfall of "verifiability", which is the bad name. "Verifiable" is based on the word "verity", or truth, and yes, we mean it as "truth that the source says X, not that X is true in itself", but all this does is create endless argument threads and unneeded discord. By focusing on "attributability", we eliminate this entire sore point by letting people know that to be included on WP, the material must be attributable to a reliable source, and if challenged or likely to be challenged, attributed. We don't focus on "truth", so we avoid having those endless philosophical discussions, while achieving the same bottom line results. Of course, if people insist on "not truth", it can go back in, but I think that with "attributability" as key, "not truth" would cause a lot of trouble with no tangible gain. Crum375 (talk) 14:34, 28 May 2010 (UTC)[reply]

I do like the new wording that makes it very clear we can use primary sources. Too many times have I seen even very long time editors remove material solely on the basis it is a primary sourced. There would be objections to the primary source even if the usage is as a source on itself or just to prove the source existed and made a statement. I also like the wording of the SelfPub section for the same reason. Alatari (talk) 06:20, 28 May 2010 (UTC)[reply]

The new ATT and NPOV

In order to stay with our pillar of NPOV, should the WP:ATT include criticism of itself and possible refutations? The ATT's heavy reliance on secondary sourced authorities could lead to the criticism that the ATT is based on the logical fallacy of appeal to authority. The requirement to evaluate sources by their fact checking and quality standards is an offsetting point. There are other criticisms of Wikipedia tied to the ATT; most directly that of Lack of Authority. Wouldn't this article be required by NPOV to state some of the ways the ATT policy is countering WP:BIAS or why it's not important to try and avoid certain criticisms? Would NPOV at least require a link in the See Other section directing a user to Criticisms of Wikipedia? Alatari (talk) 06:20, 28 May 2010 (UTC)[reply]

I think the "See also" link to Criticisms of Wikipedia may make sense. As far as this article being required to be NPOV, a policy is not an "article", so I am not sure how to apply NPOV to it. A policy "describes standards that (within the limits of common sense) all users should normally follow", so it only needs to be accurate, not "neutral" in some sense. Certainly we need to address criticisms, and if there are ways of improving the content policies we should, but I am not sure if such discussions or rationales should be embedded inside the policies proper. Perhaps there can be a separate "rationale" page for each policy, but it would probably be more appropriate as an essay. Crum375 (talk) 14:55, 28 May 2010 (UTC)[reply]

The process of policy formation is new to me. Alatari (talk) 07:41, 29 May 2010 (UTC)[reply]

Overall this is a very good rewrite Crum. Kudos. Alatari (talk) 06:20, 28 May 2010 (UTC)[reply]

Thank you. I hope it helps in some way. Crum375 (talk) 14:55, 28 May 2010 (UTC)[reply]

Teaching by examples

I'm a big fan of this style of instruction and it would be nice to have some very explicit examples in the ATT of failed attribution and successful. Showing a properly used primary sourcing and an interpreted fail usage would help clarify that section. Maybe there are some questions and subsequent discourse on the ATT talk page that would give deep insight to a new user on how to apply ATT? Alatari (talk) 07:18, 28 May 2010 (UTC)[reply]

If this were a piece of documentation we were writing for a typical organization to describe its policies, I'd agree with you wholeheartedly. But this is WP, where nothing stays constant and any IP or SPA can edit at any time, so policy pages have to be kept as short and sweet as possible. We discussed FAQ pages with examples for policies in the past, and in some cases have them, but the problem is that examples can become a magnet for controversy, and are a time sink to maintain in sync with the policy. Even the classical examples we have inside WP:SYN have caused countless talk page discussions and debates. In my opinion, essays are the best venue for examples and explanations. Crum375 (talk) 15:17, 28 May 2010 (UTC)[reply]

Rationale Q&A

I have created a rationale page for this draft policy in a Q&A format. Comments welcome: here for general ones about the draft, or on the Q&A's talk page for issues specific to the Q&A itself. Crum375 (talk) 00:11, 29 May 2010 (UTC)[reply]

Would collaboration on the editing be welcomed? Also take a look at this which might give ideas. FT2 ^{(Talk | email)} 15:50, 4 June 2010 (UTC)[reply]

I would welcome any editing which supports the basic goals which are outlined on the rationale page. The key points are to merge V and NOR into ATT with the absolute minimal changes to either letter or spirit, except as required to support the new terminology, and for general readability. The link you gave me seems to include new definitions and conceptual changes in the core policies, which is something I am trying to avoid in this proposal. Crum375 (talk) 16:30, 4 June 2010 (UTC)[reply]

Copyedits to intro

I've done some copyediting on the intro for flow and focus:

Nutshell, still brief, now explains policy. Often nutshells summarize the core principles not just the one idea. See for example nutshell on WP:NOR, WP:N, etc.
Para 1 of intro now explains core concept. Previously it went directly into advocacy on what is and isn't okay ("Wikipedia is not for..."), which is better covered simply by saying what material must comply with. (We can then detail this later)
Para 2 now explains how this impacts implicit information, which includes synthesis, juxtaposition, etc.
Para 3 now explains how non-attributed material is handled.
Para 4 is edited to be closer to the same purpose paragraph at WP:NPOV

FT2 ^{(Talk | email)} 12:42, 8 June 2010 (UTC)[reply]

Crum - nothing substantive with the exception of your substitution of the word reputable for reliable. Any rationale for that? Both mean essentially the same thing (with a slight difference). The old policy uses reliable, while this new draft uses reliable everywhere except in the lead and nutshell where you use reputable.--Mike Cline (talk) 14:37, 8 June 2010 (UTC)[reply]

Crum - This sentence: Material incapable of being sufficiently attributed is usually removed. in the third lead paragraph is problematic because the adjective sufficiently is not defined elsewhere in the policy (as far as I can tell) and is vague enough to be very unclear as to what it means. Is one reliable source OK for attribution or is that insufficient? I would rewrite this sentence thus: Material that cannot be attributed to a reliable source is usually removed. --Mike Cline (talk) 14:47, 8 June 2010 (UTC)[reply]

FT2: I have reverted your proposed version for now. There is nothing drastic that I disagree with, it's probably mostly semantics, but there are many small issues with the wording which I would prefer to get wider consensus for. The current draft wording is based almost exclusively on a combination of existing verbiage in V, NOR and ATT, and almost each word has been discussed extensively. I would prefer, if at all possible, not to introduce new verbiage unless absolutely necessary, because if there is any chance to get the community to accept this proposal, it would have to be based on a case of "format changes only". The more new verbiage we introduce, the more cans of worms we open, and the smaller the chance of success, in my view. Take as example "Material incapable of being sufficiently attributed", which Mike Cline also refers to above: I am not sure what exactly it means, it introduces new verbiage, and along with it new ambiguities. Similarly for almost every change. I would prefer at the moment to focus first on the big picture (Are we missing something which is currently in V and NOR? Are we conflicting with them in some way?), and on making sure the verbiage itself is as faithful to the originals as possible. The current WP:ATT can also be used as reference, since it is the work of many editors and reflects a lot of thought, although at the moment V and NOR are the official governing policies. I do appreciate your effort, and hope this does not discourage you, but our goal is to get this thing approved, and the community will be extremely critical about any unnecessary change from the current policies and their verbiage. Crum375 (talk) 17:47, 8 June 2010 (UTC)[reply]

It makes sense. I've worked with both approaches. In some cases I've been very careful to only reuse existing material, when it seemed necessary. But in other cases - several of them - I've taken the opportunity to rewrite a more cohesive and better policy, based on and expressing the aims of the existing ones, and that's often got consensus as well. For example my complete rewrite of WP:UP - people can read, understand what it's about, and you are no longer stuck with the issues of the past stacked up on each other. Its a lot cleaner and fixes many accumulated problems that come along now and then. But it's a complete rewrite carefully designed to fit current norms and policy, just not tied to their accumulated edits. It got high levels of approval. I've done rewrite or near-complete rewrite before on major pages including WP:NPOV, WP:ABOUT, and a dozen or so others... so I'm comfortable about major page rewrites. If they genuinely respect/reflect existing norms and will help, they often get endorsed.

I can't criticize the approach of "keep it as much as we can exactly the same only merged, to establish the principle". I understand it. But I've found that often the community is surprisingly open to a complete rewrite if done competently and respectful of existing norms. Maybe they are both valid approaches. In this case I think a rewrite might be equally useful to consider. Not to compete, more wishing to explore two approaches to a goal we both see as sensible.

So it might be useful to develop a version that's as close as possible in wording, and one with a clean start to the wording, both reflecting as much as possible the spirit of these as the community holds them. At worst nothing lost and at best perhaps some cross-fertilization will occur. Would you be agreeable, if I tried to work on a parallel, to see if you can find ways to improve it, or criticisms in it? And I'll also do what I can to help on the present version which is staying as close as it can to current wordings. FT2 ^{(Talk | email)} 18:28, 8 June 2010 (UTC)[reply]

Sure, I'd be happy to comment on your version. In this particular case, we've already been through this exercise once — this is round two. And we really need to learn from past mistakes, which included, as I noted in the rationale page, deviating in some minor ways from the exact source verbiage. If you review the last round's Oppose comments, and consider that to get this thing approved a simple majority (like the 424:354 support the last ATT version got) is insufficient, you'll see that the odds for getting new verbiage and/or novel ideas through are not good. Crum375 (talk) 19:00, 8 June 2010 (UTC)[reply]

Feedback

Crum, I sincerely apologize for the delay-- real life has done a deal on me. I do not have time to read through the full proposal, but I did read through your Q&A, and it appears that you have attempted to correct the main drawback of the old ATT proposal, which was combining WP:RS into WP:ATT. If the page conforms to your Q&A, my main objections are likely resolved. Regards, SandyGeorgia (Talk) 01:32, 17 June 2010 (UTC)[reply]

SG, thank you for your comments. I think the page conforms to the Q&A, and if it doesn't, we'll fix it. All the best for your RL deals. Crum375 (talk) 03:27, 17 June 2010 (UTC)[reply]

Government sources ??

Crum, this may not be the right place to raise this, but there is literally no mention of government published sources in the discussion of reliable sources, yet a very large amount of content in WP is being attributed to government published sources. They are no doubt reliable, but why are they excluded from the discussion? Any thoughts?--Mike Cline (talk) 15:44, 22 June 2010 (UTC)[reply]

I personally agree that most government publications are highly reliable, even for the tiniest nations. But this issue clearly belongs on the WP:SOURCES section, which is currently part of WP:V. Our goal with this draft proposal is to combine the current V and NOR into a single page, while remaining as faithful as possible to their current versions. See also "suggested improvements to the current policies" in the rationale page. BTW, an advantage of the new ATT is that WP:SOURCES and WP:PSTS are now side by side, and for this specific issue, where you have highly reliable sources which are most often primary, it's useful to have the relevant policy sections in one place. Crum375 (talk) 16:15, 22 June 2010 (UTC)[reply]

"Parties" are a closely related concept to government documents. Government documents are often highly reliable, but since the government is involved in almost anything, the government is seldom a third party. Third party ≠ independent. I cannot support any proposal that claims or implies otherwise. You will probably complain that the current policies already claim third party = independent, and you are just trying to eliminate overlap and poor terminology in current policy. Well, "third party" is poor terminology and I will not support anything that mentions "third party" unless the meaning is confined to its proper realm, contract law. Jc3s5h (talk) 19:37, 22 June 2010 (UTC)[reply]

As described in the rationale page, the goal of this ATT draft proposal is to not introduce any conceptual or verbiage changes to the existing policies, unless absolutely necessary. Do you see anything in the draft that deviates from the existing policies in letter or in spirit? Crum375 (talk) 22:08, 22 June 2010 (UTC)[reply]

Also, about your specific point, I believe that, as I noted above, government reports are most often primary sources, and normally cannot used where secondary sources are required. And I agree with you that typically the government is directly involved in some way with what it's reporting on, which is the reason its documents are almost always primary sources. But having said that, our goal here is to simply combine the existing V and NOR into one page, and not introduce any conceptual or non-essential verbiage changes. Crum375 (talk) 22:29, 22 June 2010 (UTC)[reply]

I feel that saying "third-party" can cause people to go off on non-productive tangents about the organizational relationship between a party mentioned in an article and an author (or publisher) of a source. The real issue is independence and being disinterested, not incidental legal relationships. An additional point is when an article about a topic, such as climate change, rather than about people and organizations. Since concepts have no first or second parties, the term "third party" has no clear meaning and invites useless wikilawyering.

I consider these tangents to be of the same character as writing "no original research" when we really mean "not attributable to a reliable source" and both instances of poor terminology should be corrected at the same time.

I don't necessarily disagree with you, although for me personally "third party source" and "independent source" mean the same thing. But this is the current terminology, and my fear is that if we try to change things while combining V and NOR, we'll end up collecting a lot of minor changes, which taken together will become an obstacle to approval by the community. If you read the Oppose comments in the old ATT poll, you'll notice that a lot of them pick on the tiniest things, and I think that even with no change whatsoever it will be a tough challenge to get a wide enough consensus to pass, even without any changes. This is why I feel we need to resist the temptation to make changes now, and address them after (or if) the combined version is accepted. Crum375 (talk) 23:55, 22 June 2010 (UTC)[reply]

Feedback: Nice Work!

I think that it is very clear that these two policies should be combined, this and the rationale page represent an EXCELLENT job of doing so. Even though we often disagree (I think that some changes in the policies are needed) such is not relevant here. This combining needs to be done, and you have done an excellent job at drafting a combined policy and it's rationale.

Your "this shall be just combining, not changing-while-combining" approach is essential to stick to. If we started considering doing combining and changes at the same time, this proposal would die under it's own weight.

Was Slim also working on one of these? If so that should get worked out.

If you are looking to build up a head of steam before officially proposing this, I think that it needs a few more/ more substantial mentions and a little more time on the wp:ver policy talk page. We don't want this to seem like a surprise from wp:nor folks. North8000 (talk) 20:57, 22 June 2010 (UTC)[reply]

Concur. You've seen this through so far into an excellent rendition of current policy. I trust we'll find a way to generate wider consensus. Thanks for your efforts and you'll have my continued support in seeing this to a logical conclusion.--Mike Cline (talk) 21:20, 22 June 2010 (UTC)[reply]

Thank you both, North and Mike, that's very encouraging to hear. My plan is to try to gradually build up support for this, while addressing any concerns. I fully agree with North that any attempt to "tweak" the policies while combining them is doomed to failure. I think people are (understandably) jittery about any changes, and it will be hard enough to just get the basic structure and format changes approved, so any extra wording changes could add major hurdles and reduce the chances for wide-spread acceptance. If anybody sees anything in the current policies, either letter or spirit, which is not reflected in the combined version, or anything else that could be done to improve the draft without deviating from the current V and NOR, please let me know. Crum375 (talk) 22:21, 22 June 2010 (UTC)[reply]

Your effort deserves the support of the community. Excellent! PYRRHON talk 04:12, 23 June 2010 (UTC)[reply]

Thank you! Much appreciated. :) Crum375 (talk) 04:34, 23 June 2010 (UTC)[reply]

Deal with WP:NOR first and leave WP:V intact and separate for now

This is right direction, but I fear it is too big a step for the community. I don't think we should try to merge WP:V and WP:NOR at the same time as altering the WP:NOR problematic terminology.

The terminology problems of WP:V are not so bad. I don't think people really get into knots over verifiability=veritiy=truth. I think most people quickly get it that verifiability is about verifying that the content in really in a reliable source. The WP:NOR terminology is slightly worse. The meaning of the term is not obvious, and wikipedia-definitions have gone beyond reasonable deductions of what the term means.

I suggest a smaller step. Transform WP:NOR alone into WP:ATT. Minimise the overlap between WP:V and WP:ATT to a degree, but this is a lesser problem for a later date. In WP:ATT, define WP:NOR as a significant but special case of WP:ATT. There are many ways to make material that fails to be attributable without violating a reasonable expectation of the meaning of WP:NOR. WT:NOR regulars have tended to push their credibility of their use of the vernacular in arguing otherwise.

I hope that I am not being read as a naysayer. The merger is probably a good idea. "Attributable" is obviously a better word for what we want in our material. However, over-ambition can kill good ideas, and there is no deadline. --SmokeyJoe (talk) 01:50, 23 June 2010 (UTC)[reply]

SJ, as I see it, it's really all a single issue. We need a single clear unified policy which tells us that all material on Wikipedia must be attributable to a reliable source. The rest is just details. Of course that may be where the devil is, but a good policy should be able to do a reasonable job presenting and clarifying those details. To split up attributability into two parts would destroy its core message, in my view, which is the key concept that everything must be attributable. By splitting it, we'll just perpetuate the current confusion, which I believe is at least partly due to the current split, with its overlap and poor terminology, and the effort will be worthless. As I see it, either we can get this unified attributability accepted by the community, or we should find other things to do with our time. But I appreciate your candid input. Crum375 (talk) 02:05, 23 June 2010 (UTC)[reply]

Hopefully, I err on the side of pessimism. I still say that we are skipping a stepping stone on this path, but you may be right, in that putting something called "WP:ATT" on that stepping stone may be bad, as that thing is not what people expect WP:ATT to be. Perhaps if we created a new page to show what WP:NOR would look like if adjusted to the language of WP:ATT. Maybe...? --SmokeyJoe (talk) 02:29, 23 June 2010 (UTC)[reply]

There are two issues with this approach, at least. First, NOR today consists of PSTS, SYN, and assorted verbiage. In the proposed ATT draft, PSTS goes to "Sources" and SYN to "Content". This important organizational step, moving PSTS to its natural home with sources, is one of the key features of the combined draft. I therefore don't see a logical stepping stone for this, since the "Sources" section in the draft ATT is mostly where the current V goes.

Second issue is more general: I think the odds for success are greater by presenting to the community a single, unified, coherent combined version, than a piecemeal road map, built on awkwardly defined, overlapping pieces, which will only add to the current confusion.

As bottom line, I think we need to create a single, clear and coherent proposal, and let the community decide if they find it attractive. To try to mealy-mouth our way to a solution will not work, in my opinion. Crum375 (talk) 02:45, 23 June 2010 (UTC)[reply]

I originally also suggested the "baby step" approach, but I think Crum's current idea is better. A clean merger that does not change anything except get rid of the OR term, and renaming ver to att. That should avoid controversy except from those who not comfortable with the merger or the renaming. Trying to make an (in some ways) big change by piecemeal editing is likely to have many more potential pitfalls. North8000 (talk) 11:40, 23 June 2010 (UTC)[reply]

I like Crum's answers. I am a little hurt by the suggestion that I suggested to mealy-mouth a way to the goal, but I will try to live with this. I definitely support the goal here. --SmokeyJoe (talk) 04:18, 25 June 2010 (UTC)[reply]

Joe, thank you for your kind words and your support. I am sorry if I expressed myself poorly, and apologize for any unintentional offense. All I was trying to say is that I prefer to go through this policy update process only once, in a single step. As you know, getting any significant policy change broadly approved by the community is very difficult, because many people are naturally conservative and don't like changes, on principle. In fact, I myself am generally like that. Therefore, my plan here is to proceed slowly, and to try to get everyone on board before we even hold a poll, by addressing all their concerns ahead of time. This is why I prefer to keep this draft in user space for now, until we can get sufficient grass root support to move to a more public location. Thanks again. Crum375 (talk) 05:13, 25 June 2010 (UTC)[reply]

No offense was taken. Mild humour intended, plus acknowledging the probable truth of your words. --SmokeyJoe (talk) 23:36, 25 June 2010 (UTC)[reply]

Thoughts

I don't oppose a merge of WP:V and WP:OR, but I have some quibbles with details. We should call this policy "Use reliable sources" as that's what it says over and over again in different ways, and that's what we want people to do!
There are a couple of key catchphrases in explaining Wikipedia to newbies: "Verifiability, not truth" and "No original research". This does away with both of those! I don't see why the first catchphrase needs to vanish entirely as it encapsulates our approach succinctly.
We need some kind of succinct statement to say that Wikipedia is not the place for previously unpublished ideas or observations to replace "no original research", if this phrase is indeed too ambiguous: what about "No unpublished material", "No unpublished content", "No unpublished facts, ideas, or opinions", or "Nothing unpublished" (WP:UNPUB)?
I don't see that anyone is confused by "verifiability" as a phrase and it's actually a more precise description of what we require than "attribution": we need to be able to verify that the source says what an editors asserts it says; attribution alone isn't enough.
Btw, what happens to Wikipedia:Identifying reliable sources? I didn't see that linked to anywhere prominent. Fences&Windows 15:52, 23 June 2010 (UTC)[reply]

(ec) Thank you for your detailed comments, Fences. Let me start with the first one: "use reliable sources", though important, is not the key to WP content policies. In fact, we may be perfectly policy-compliant not to use any source at all in some situations, as long as the material is attributable. It is the attributability of all content material that is the key, everything else is details. Regarding the "verifiability, not truth" catchphrase, see my reply above to this point. You say that people are not confused by verifiability, but there have been many long talk page threads indicating the opposite, and one key goal of this draft is to eliminate potential sources of confusion. Regarding IRS, the highlights of that guideline has been mostly subsumed into the WP:SOURCES section over the last couple of years. In general it's not a good thing to point from a policy into a guideline, but the other way around. As it is, there are constant discrepancies between them that need to be addressed. If there is anything crucial on IRS which is still not in WP:SOURCES, it should be added to WP:SOURCES (in V for now), which is the controlling policy. Regarding your point that "attribution alone isn't enough", you are correct. Per WP:BURDEN, what we need is to prove to the satisfaction of the challenging editor that the material in question is supported by a reliable source. WP:BURDEN tells us that "the source should be cited clearly and precisely, with page numbers where appropriate, and must clearly support the material as presented in the article." That is the key: all content material must be attributable to a reliable source, and when challenged, we need to attribute it to a reliable source, and must prove to the challenger that the attributed source "clearly supports" the material in question. The word "verifiability" adds confusion, because many newcomers incorrectly (but justifiably linguistically) assume it means we need to "verify the truth" of the material. This draft attempts to avoid this confusing terminology by using the simpler, more direct term "attributability", which leaves no room for doubt. Crum375 (talk) 16:51, 23 June 2010 (UTC)[reply]

To further my point about "attribution", I shall tell a story:

User:Professor Joseph Bloggs III signs up to Wikipedia, and he says he's a world expert on the study of cruftiness. He confirms with OTRS that he is indeed THE Prof. Joseph Bloggs III, and he begins to edit our article on Cruftiness, but he uses no sources. Quelle horreur! "Prof. Bloggs, with respect," say the Wikipedians, "we need all fact and opinion on Wikipedia to be "attributable"." Prof. Bloggs replies that "It is attributable: it is attributable to me, and I am an expert." The Wikipedians respond by opening and closing their mouths several times, before quietly sneaking off to edit the policy that uses such an ambiguous term.

"Attribution" is not a good term to use. Fences&Windows 16:36, 23 June 2010 (UTC)[reply]

Prof. Bloggs would be correct, so long as the material he adds to Wikipedia is attributable to a reliable published source, including his own publications which meet our reliability requirements. If he feels there is little chance they would be challenged, and there are no quotes, he may elect to skip the attribution, so long as he knows he can attribute it at any time. Once challenged, he must attribute the material to a reliable published source (including his own), and must prove to the satisfaction of the challengers that the source clearly supports the material in question. On Wikipedia we allow editors to rely on their own reliably published sources, as long as those sources meet our reliability requirements, and an attribution to a reliable published source is provided when challenged, likely to be challenged, or quoted. Crum375 (talk) 17:02, 23 June 2010 (UTC)[reply]

Let's say for the sake of argument that it's unpublished material, so my point stands. You can attribute a previously unpublished idea to someone, e.g. <ref>Personal communication with subject of the article.</ref> (such an attribution is common in scholarly literature). "Attribution" is even more prone to misinterpretation than you claim verifiability is. If you have to append "to a reliable source" for it to match Wikipedia's usage, you're showing that on its own it is not a useful term. Fences&Windows 19:12, 23 June 2010 (UTC)[reply]

You can have just as many problems with a published source, for example a self-published one, or a misused one. The point is that when we say "sources", unless otherwise qualified, we mean "published by reliable sources and conforming to WP:SOURCES". So when we require "attribution to a reliable source", it's short-hand for "attributed to a published reliable source, conforming to WP:SOURCES." Crum375 (talk) 19:51, 23 June 2010 (UTC)[reply]

You say that ""use reliable sources", though important, is not the key to WP content policies." Really? This page begins: "Wikipedia is a compendium of knowledge published by reliable sources, so all material must be attributable to a reliable published source to be eligible for inclusion." Now, I understand the -able/-ed distinction, but the key to WP:V and WP:ATT is using reliable sources, whether they are explicitly cited or not. Using reliable sources has become the cornerstone of how we write Wikipedia, it is how we check facts and opinions, decide WP:NPOV (one of the five pillars), defend against BLPROD, etc. "Use reliable sources" is the most basic of advice for a newcomer to Wikipedia, and is basically the core rule (aside from WP:IAR and WP:NOT). Fences&Windows 19:24, 23 June 2010 (UTC)[reply]

Fences, I am not saying reliable sources are unimportant; in fact, they are crucial to what we do here. But they are not the key to the content policies. The key is attributability, i.e. that everything on Wikipedia, explicit or implicit, must be directly and clearly attributable to a reliable source. There is no question that reliable sources are needed, but that in itself is not sufficient, because, for example, people might (and often do) synthesize new information by combining two pieces from individually reliable sources, violating WP:SYN. Or another common example is editors providing highly reliable sources, which on close examination don't quite support the material in question. So in summary, the key, or the top level view, is attributability — everything must be attributable to a reliable source, which directly supports it. Crum375 (talk) 20:05, 23 June 2010 (UTC)[reply]

'Attribution' as a word outside Wikipedia means noting who wrote or said something, not providing evidence that they wrote or said this. Journalists 'attribute' comments to people without providing recordings of interviews or signed statements. People's common understanding of 'Attribution' differs subtly but importantly from the meaning you are assigning to the word within Wikipedia: this is not good if you want our policies to be immediately understood by newcomers. Fences&Windows 16:26, 25 June 2010 (UTC)[reply]

I agree with some of what you said. If we just "attribute" something to someone, we are not (necessarily) providing any evidence for it. This is why on Wikipedia, and specifically in this proposed policy, we say that when attribution is required (i.e. for challenged, likely to be challenged or quoted), "Editors should provide attribution to a reliable source using an inline citation...The burden of evidence lies with the editor wishing to add or retain the material." In other words, it's not enough to provide the attribution, but the editor adding the material must convince the challenging editor that the source supports it. We explain the "burden of evidence" in the Burden of evidence sub-section of Key principles (bold added): "The burden of evidence lies with the editor who adds or restores material. All quotations and any material challenged or likely to be challenged must be attributed to a reliable, published source using an inline citation. The source should be cited clearly and precisely, with page numbers where appropriate, and must clearly support the material as presented in the article." So you are correct, in that attribution alone is insufficient, as the person providing the attribution must be able to convince the challenging editor that the source actually supports the claim, per WP:BURDEN. Crum375 (talk) 23:55, 25 June 2010 (UTC)[reply]

re: Citing yourself

Hi, first of all, good effort, reads easily and gives lots of arguments.

Now my comment:

The section :Citing yourself; should in my view also include a reference to media sources the topic has more or less complete control of. For example

Silvio Berlusconi is owner of the major television stations in Italy, and as prime minister has almost absolute control over Rai Uno. He has actually sacked editors for showing him in bad light.
Rupert Murdoch owns a lot of media, that are giving the POV of Murdoch in many cases (e.g. Fox Broadcasting Company
Fethullah Gulen owns, through his movement, the influential Zaman (newspaper)

In all three cases, the media sources, that may otherwise be respected and important media (television, newspaper etc) will always publish positively interpretations of whatever their controlling owner does. Therefore, I would like to include any sources which are (largely) controlled by a person to be labelled as (by definition) not reliable on any topics related to the controlling person; following WP:COI. Maybe this is a bit far fetched and I should take this somewhere but I think COI would cover this and since you are compiling this storyline.... Cheers. Arnoutf (talk) 21:12, 23 June 2010 (UTC)[reply]

Hi Arnoutf, thanks for the kind words. As far as the specific issue you raise, you may want to see WP:SELFCITING, which is part of the current WP:NOR policy. Our concept here, per the rationale page (see "suggested improvements", currently item #15), is to reflect the official WP:V and WP:NOR policies faithfully, with the identical verbiage wherever possible. So in this case, you may want to raise the issue on WT:NOR. If a tweak gets accepted and remains stable, we'll incorporate it here too. Thanks again, Crum375 (talk) 21:26, 23 June 2010 (UTC)[reply]

Is this Proposal Incubating or is it On-Ice?

Is this Incubating or is it On-Ice?

If "incubating", I would propose taking the next baby step forward.

If "on ice" then I think it's time to start eliminating some of the overlaps between wp:ver and wp:nor, and fixing one gap in wp:ver with material from wp:nor. North8000 (talk) 11:18, 5 August 2010 (UTC)[reply]

I would say it's waiting more input from more users. There are many experienced editors who have yet to comment here, and I think it makes sense to get their views before proceeding. Crum375 (talk) 13:05, 5 August 2010 (UTC)[reply]

The last input was 6 weeks ago. Do you want me to ask for input at wp:ver and wp:nor? I'd be happy to. North8000 (talk) 23:03, 6 August 2010 (UTC)[reply]

Sure, any additional input would be helpful. Thanks. Crum375 (talk) 23:09, 6 August 2010 (UTC)[reply]

Crum your fans are calling at wp:nor talk . North8000 (talk) 23:48, 26 August 2010 (UTC)[reply]