Wikipedia talk:Attribution

The project page associated with this discussion page is an official policy on Wikipedia. It has wide acceptance among editors and is considered a standard that all users should follow. Before you update the page, make sure that any changes you make really do reflect consensus.

Editors, please note:

After four months of discussion at Wikipedia:Attribution, editors at Wikipedia talk:Attribution have agreed on a means of merging Wikipedia:Verifiability with Wikipedia:No original research, while also streamlining Wikipedia:Reliable sources into a simpler FAQ at WP:ATT/FAQ.

There are no policy innovations suggested: WP:ATT is intended be a more cohesive version of the core content policies with which the Wikipedia community is already familiar.

"in articles about themselves"

The wording keeps getting changed back to "in articles about themselves", but this isn't the actual criteria. This wording is both too broad and too restrictive for neutrality. Questionable sources should not be used for concrete statements of fact, but they should be used for statements about themselves or their authors. In other words, you can't say "the earth is flat", and use someone's blog as a reference for it, but you can say "blog writer x says the world is flat" and use the blog as a reference. It has nothing whatsoever to do with being in an article about the blog or about its author. — Omegatron 07:36, 9 March 2007 (UTC)

What makes you think a random blog passes the test for use as a reliable source in the first place ? Wjhonson 08:08, 9 March 2007 (UTC)

It is a reliable source in an article about the blog or its author. But why such an article would claim the Earth is flat is beyond me... JulesH 10:31, 9 March 2007 (UTC)

"It has nothing whatsoever to do with being in an article about the blog or about its author." Yes it does, because such a statement would have no place on any article but one about the blog or its author. Marskell 10:37, 9 March 2007 (UTC)

I have an issue with the idea that any fact is not contentious. The current article makes it sound like we can use an anime fan webiste as a citation for anime fans being predominantly male. Anything is contentious once somebody says it might not be true. Lotusduck 01:07, 14 March 2007 (UTC)

The flat earth example isn't very good. For one thing, you don't need a blog for that claim. There are much better secondary sources for the existence of a tiny worldwide minority who support that idea, e.g., the Flat Earth Society.

"Flat earth" is actually an example of a view which has too few supporters to be mentioned or even linked to in astronomy and geography articles. Okay, maybe one quick mention that people may have believed in a flat earth in pre-literate times.

But there's not much of a controversy over it these days. More important is when there is a political controversy over a legal or scientific matter, and when the proportion of partisans is roughly equal (like 40-60 or 50-50) or when the minority is loud or obnoxious. --Uncle Ed 00:32, 19 March 2007 (UTC)

Cloning and stem-cell research come to mind. — SMcCandlish [talk] [contrib] ツ 01:38, 19 March 2007 (UTC)

Primary source

See: Archive 11: Primary source --Philip Baird Shearer 12:14, 9 March 2007 (UTC)

I don't see your point. There is a definition of what a primary source is, and a number of good examples too. JulesH 12:21, 9 March 2007 (UTC)

The change I am making is to put similar information after the definition, instead of using the term before it is defined. Also it is not edits to text that are covered by this policy document but rather "All material in Wikipedia" (the text once the edits have been made). --Philip Baird Shearer 12:38, 9 March 2007 (UTC)

I don't see a problem with using a term and then defining it in the next sentence. This is fairly standard practice and helps the document flow better. I'm also not sure why you're picking solely on primary sources when the exact same arrangement is also used for secondary sources, 'original research', 'reliable sources', 'questionable source', and 'self-published source'. I do agree with you that the reference to 'edits' is perhaps wrong, and should probably be changed to 'articles'. JulesH 12:43, 9 March 2007 (UTC)

I have not yet altered secondary sources etc., because incremental changes are better than large changes. You may think that it helps the document flow better, but the structure of this pivotal policy document should be close to a well written legal document, and in those definitions always come first to avoid ambiguity in the text. After all the reason for combing WP:V etc into one was to clarify things which were already there but in several documents and it helps if this document should be constructed in such a way as to reduce ambiguity still further. One way to do this is to make sure that it is structured logically.

Also I have problems with "The Bible cannot be used as a source for the claim that Jesus advocated eye removal (Matthew 18:9, Mark 9:47) for his followers, because theologians differ as to how these passages should be interpreted." Why pick out the Bible for consideration why not the Koran or any other religious book? The current wording implys that the Bible holds a special position inside Wikipedia. --Philip Baird Shearer 13:04, 9 March 2007 (UTC)

The bible is the most widely distributed primary source material in the English language. This is the English Wikipedia. The Koran is not generally translated into English, so few of the English-speaking readers of this web site will be familiar with its content. The reasoning behing this choice is obvious.

I also disagree that the structure of a wikipedia policy should mimic that of a legal document. Most people find reading legal documents difficult and tedious; wikipedia is not a beurocracy and we want to encourage people to understand the policies rather than needing lawyers to interpret them for them. The policy should therefore follow the structure of documents designed to inform, i.e. educational texts. The style of using a word then explaining its meaning is common in such texts. JulesH 14:17, 9 March 2007 (UTC)

Choosing the Bible may be in your opinion "obvious" but it is not in my opinion an example of systemic bias and probably breaks the WP:NPOV policy. If this needs to be mentioned it should be in the FAQ not in this policy article.

How do you know that most people find reading a legal document difficult and tedious? I think it depends on the document. In the past when solicitors were paid by the word I would have agreed with you, but modern contracts drawn up with a consideration for plain English are much easier to understand. I put it to you that if this article is structured in such a way as to minimise inconsistencies, it is less likely to misunderstood. --Philip Baird Shearer 16:34, 9 March 2007 (UTC)

And I suggest that if we introduce the reason why the reader of the policy should care about a concept before describing the meaning of the concept, it is more likely to be read. What precise ambiguity is introduced by using the phrase "primary source" in the sentence before it is defined?

And, what other source can we use for an example that a significant proportion of our readers will be familiar with? JulesH 21:12, 9 March 2007 (UTC)

One comment -- generally speaking, when writing about research or scholarly work, particularly recently published research that's reported in news media, given a choice between citing the original scholarly publication and citing a newspaper or other popular media account of it, citing the actual work should be prefered, and one should basically always cite the actual work where possible. Newspaper accounts on new studies etc. can sometimes distort and create a sort of telephone effect, citing the actual work both makes clear what the work is and enables some future expert to tell if it's been reliably described. In these cases, suggest that primary sources should be preferec to secondary ones even though the newspaper accounts will be in plainer English and easier to understand. One can always cite both. Best, --Shirahadasha 06:52, 11 March 2007 (UTC)

Getting back to the Bible example: the reference is to Jesus saying that if your eye or hand causes you to commit a sin, it's better to remove it than go to hell (paraphrasing for clarity here). Check out Mark 9:42-47 for the full context.

But the reason this is a good example has nothing to do with "Western" or "Christian" preference for the New Testament. It's rather a matter of interpretation. Traditionally, Christians have seen Jesus as using vivid physical to make a point. I cannot recall anyone, even a Fundamentalist, ever saying, "If you feel tempted to commit a sin, just commit suicide instead." And the example in Wikipedia:Attribution cautions editors against taking a scriptural passage literally, when it is mostly interpreted as a metaphor or analogy or 'rhetoric'.

Come on, you're killing me with this nitpicking. ('Killing me' is rhetoric, and 'nitpicking' is a metaphor.) --Uncle Ed 00:42, 19 March 2007 (UTC)

Burden of evidence

The introduction to this statement of policy mentions the "burden of evidence". I've never heard of the burden of evidence. Ought it not read "burden of proof"? DavidCBryant 15:33, 9 March 2007 (UTC)

What's really required is some evidence that backs up the editor's claims, not necessarily definitive proof. Evidence is really a minimum requirement since without any evidence there's certainly not proof. ChazBeckett 15:43, 9 March 2007 (UTC)

The statement means that the burden of providing evidence to back a claim (ie a citation to a reliable source) is upon the person wishing to add the claim to the article. Blueboar 16:55, 9 March 2007 (UTC)

Exactly, and not to be confused with the burden of evidence about the truth of the cited assertion. ≈ jossi ≈ (talk) 16:58, 9 March 2007 (UTC)

Indeed. If we said "the burden proof" people would think in terms of the truth of the assertion ... by saying "burden of evidence" we make it clearer that we are talking about who has to add sources. But perhaps "burden of providing evidence" would be clearer yet? Blueboar 17:12, 9 March 2007 (UTC)

question about application of Wikipedia:Attribution

What is the rule concerning a user who removes text because of Original Research of their own? For example, a use who removes an argument because they think its false based on their own original synthesis of other sources?--Urthogie 17:18, 9 March 2007 (UTC)

Since we should not have argument based on an editor's original synthesis in the first place, the editor not only may, but is encouraged to remove it. If you like the argument and wish to keep it in the article, go look for an independant secondary source that states it. Blueboar 18:23, 9 March 2007 (UTC)

I think you misunderstand my question. I'm saying when the removal itself is based on the removers OR.--Urthogie 19:02, 9 March 2007 (UTC)

This is not prohibited. Neither, however, would restoring the information be. —The preceding unsigned comment was added by JulesH (talk • contribs) 21:20, 9 March 2007 (UTC).

I would say it is prohibited. Jayjg ^(talk) 22:32, 9 March 2007 (UTC)

JayJg - Which is prohibited, removing or restoring? Blueboar 22:47, 9 March 2007 (UTC)

Of course it's not policy violation to restore non-OR (unless its 3RR). My question, though, is if its a policy violation to remove something because you find it to not be true based on your own obvious original research. And this is assuming everyone editing agrees with you that X isn't true after you show your evidence.--Urthogie 23:24, 9 March 2007 (UTC)

It is prohibited to remove relevant material sourced to reliable sources because you have done some original research and decided the reliable source is actually wrong. Jayjg ^(talk) 01:51, 11 March 2007 (UTC)

(unindenting) I'd say it depends. If by "original research" we are talking about an experiment, I'd say no: if you have come up with a rigorous experiment that contradicts received opinion, the best thing to do would be to see how you can get your results published in the appropriate periodical. If by "original research" we are talking about a secondary source that claims that source X had a given passage in it, & you have read source X & found that it does not -- I'd say that is possible grounds to remove that use of the secondary source. I'd try to find one or more Wikipedians who know the subject, show them your findings, & solicit their input first, though; there may be good reasons why this happened. -- llywrch 23:59, 9 March 2007 (UTC)

I think Llywrch is addressing the issue. Sometimes editors reading about "original research" get it confused with "source-based research". Original research must assert a new fact, doing research to find that person X asserted a new fact is not itself "original research", but rather "source-based research". So perhaps the original poster can give us the specific example in this case? Wjhonson 00:48, 12 March 2007 (UTC)

Another question, I don't see the new policy mentioning tertiary sources. Are they considred to be reliable sources? --Brian Wiseman 00:35, 10 March 2007 (UTC)

WP:ATTFAQ says that "[t]raditionally published encyclopedias are reliable sources". All the same, unsigned encyclopedia articles are hardly desirable sources. Angus McLellan (Talk) 01:13, 10 March 2007 (UTC)

Effects on External links

Via the WP:EL#Links normally to be avoided #2 guideline, this policy page now applies to external links (normally). The nutshell states that this policy applies to material in Wikipedia. Was this intended? AndroidCat 17:50, 9 March 2007 (UTC)

Your question really should be asked at the EL page... if the original intent at EL was that WP:V, WP:NOR and/or WP:RS were to apply to External Links, then this page (or the relevant sections) should apply as well, since this policy is a continuation of those Policies. Blueboar 18:29, 9 March 2007 (UTC)

External links are material in Wikipedia, but that doesn't mean the content of the sites we link to must adhere to ATT. SlimVirgin ^(talk) 19:15, 9 March 2007 (UTC)

The guideline in question reads: Except for a link to a page that is the subject of the article or an official page of the article subject—and not prohibited by restrictions on linking—one should avoid: ...

...Any site containing factually inaccurate material or unverifiable research.

It then gave a wiki-link to the old WP:RS guideline. I think the purpose was to help the editor define what is meant by "verifiable resarch". If that is the case, then the intent is the same when linking to this Policy. If they mean something else, or find that their intent is not met within this page, then it is up to the folks at EL to re-word their page. Blueboar 19:30, 9 March 2007 (UTC)

Thanks everyone, that answers my question and I'll take it up in EL if I absolutely have to. :) AndroidCat 22:54, 10 March 2007 (UTC)

Weak sources or sources used outside context

In the article Norouz as celebrated by Kurds, this source has been used:

"Rasbridge, Lance Andrew; Kemp, Charles (2004). Refugee and Immigrant Health: A Handbook for Health Professionals. Cambridge University Press, pg. 236. ISBN 0521535603."

To support a statement regarding legends behind a spring festival. This book is obviously focused on something completley different, not only to the festival, but to the ethnic group we are discussing here. I believe there was just a note regarding this festival to explain hazards probably regarding injuries caused by jumping over a fire (a tradition of this festival) or something similar.

I wanted to know whether this can be used as a primary source to support a statement and whether it can be called a "weak" source. It must be appreciated that the statement has another reference too so I am just using this as an example and do not object to the statement that the reference has been used for here. Regards, --Rayis 16:23, 10 March 2007 (UTC)

The source should be fine if the statement within it is reliable and supports the claim. The need for a claim to be a primary subject in the souce is critical to establishing notability but doesn't make the source weak. Of course, if there is completely contradictory information from a more reliable source on the matter, such as a scholar of Kurdish issues, the scholar should accordingly receive more weight. The Behnam 19:46, 10 March 2007 (UTC)

Non-scholarly sources

In a recent AFD discussion where someone was complaining of the sort of sources used in an article on a fictional subject, I had attempted to provide them a link to the section in the old WP:RS that discussed non-scholarly sources. When i discovered the recent merger and that the section was no longer included i looked and discovered that the section had been removed from the original RS on Feb. 24th without any discussion. I am restoring the section to the main page here, as it was removed without consensus in the first place. If anyone would like to discuss changes to the section I invite them to do it here. -ΖαππερΝαππερ ^Babel_Alexandria 16:53, 10 March 2007 (UTC)

It is in the WP:ATT/FAQ. ≈ jossi ≈ (talk) 18:29, 10 March 2007 (UTC)

Individual sections need tagging

I'm noticing that there are shortcut links to bring readers/editors directly to core principals (like NOR). This is fine but unfortunately it can be a bit disorienting to suddenly "teleport" via a hyperlink to a section and not immediately understand what it is pertains to. I think it would be sensible that each shortcutted section be tagged as being representative of policy so that individuals are not obliged to review the entire page (and in particular the top) to understand they are looking at Wikipedia policy. I would like to edit these features in. Is there anyone who disagrees with this idea? (→Netscott) 17:47, 10 March 2007 (UTC)

I'm not sure I understand what you mean. Could you describe exactly what you intend to do? JulesH 22:17, 10 March 2007 (UTC)

Sure, it is not uncommon to find links like this: WP:NOR as we edit Wikipedia. Obviously NOR is jargon for No Original Research but for editors not familar with the the term it's just a TLA. So when one clicks on a WP:NOR link they are instantly whisked away to the Wikipedia:Attribution#No original research section of WP:ATT. For such an inexperienced person to understand that what they are reading is policy they are (currently) obliged to scroll up to the top of the page to see the {{Policy}} tag. What I am proposing is to label each section with a mini tag so that when a user is brought to that section of the policy page they will instantly recognize that what they are reading is policy. (→Netscott) 01:38, 11 March 2007 (UTC)

Netscott, I added a new "policy section" template to the NOR and RS sections in line with your concerns. My only worry is that it might lead newbies to think that somehow the other sections aren't official policy. If others want to remove the new template, I'm fine either way. SlimVirgin ^(talk) 02:27, 11 March 2007 (UTC)

I think that adding a tag per section could get confusing, and create a bad precedent. The main problem I see, which SV noted above, is that putting the tag on one section necessitates doing it for all, which would create clutter and confusion, since we also have one for the entire page. I think we need to assume some minimal competence and intelligence on the part of our newer editors - it doesn't require a rocket scientist to scroll up to the top and get some context. Crum375 02:44, 11 March 2007 (UTC)

Well the section tag would need better wording, mainly that each section represents a part of an entire policy page. I was thinking of coming up with some slimline templates for such a purpose in line with what I am talking about here. (→Netscott) 02:48, 11 March 2007 (UTC)

My concerns are precedent, confusion and clutter. I can see how once we have these sections tagged as 'policy', every section of every policy would require one, or else we would have inconsistency and confusion. Also, we already have way too many templates - this is very disorienting to newer (and some older) editors. I think we need less, not more, templates. I still think that one 'policy' template at the top certifies the entire page - adding section tags diminishes the role of the main one, IMO. Crum375 02:53, 11 March 2007 (UTC)

There doesn't really even need to be tags... it could be as simple as a policy symbol much like the {{sprotected2}} symbol off to the right that'd be clickable and take one to the top of the page or something along those lines. (→Netscott) 03:04, 11 March 2007 (UTC)

Well, smaller and less intrusive is better clutter-wise, but it may not address the precedent and confusion side, as we would need it for every single policy section site-wide to be consistent. Also, if they are small icons, then their utility for the newly-landed editors who don't scroll up would be reduced also, since they may miss or misinterpret the icons. Crum375 03:09, 11 March 2007 (UTC)

I just made this. Would something like it work?

≈ jossi ≈ (talk) 03:31, 11 March 2007 (UTC)

Let's see:

That has potential. :-) (→Netscott) 03:33, 11 March 2007 (UTC)

I could attempt to make a version in SVG format so that it scales better in smaller sizes. Will take some effort to do that, so I will wait for aditional comments and assess if there is some basic support for such a "seal". ≈ jossi ≈ (talk) 03:37, 11 March 2007 (UTC)

My response was only half serious. I like the logo though. What I was envisioning was just to use the standard green check (and be clickable to the top of the page) something like this: (→Netscott) 03:41, 11 March 2007 (UTC)

I like the logo a lot. It shouldn't be too small or newbies will likely miss it. SlimVirgin ^(talk) 03:41, 11 March 2007 (UTC)

I would work a bit on making a logo in SVG format (vector) that will scale well and would be readable at smaller sizes. It may be useful. ≈ jossi ≈ (talk) 03:59, 11 March 2007 (UTC)

It does look nice, but again my question is: are we going to put it on every section of every policy site-wide? If not, which criteria do we use to decide which section gets it and which doesn't? And what happens to the ones that don't, do they then seem less important? And overall does this reduce confusion or add confusion and/or inconsistency? I can see the motivation for this, but my concern is that the solution could add more problems, if applied inconsistently. I can see using your nice icon (or similar one) at the top of every policy page, though. Crum375 03:42, 11 March 2007 (UTC)

Well there is a bit of a precedent actually: WP:FU with the policy section. I don't forsee a problem with adding small policy indicators. I'm really only proposing this as ATT has brought together under one roof all of these former policy pages that had a bit more "punch" and I'm just trying to reestablish some of that lost "punch". (→Netscott) 03:46, 11 March 2007 (UTC)

The WP:FU example is exactly what I am worried about. It seems that in that case there is a policy embedded inside a guideline. So if we were to add an icon or tag on a section here, we'd have possible confusion about whether the entire page is a policy or not. If not, then why is it needed per section? And why not for every section of every policy site wide? I believe in simplicity and consistency. I fully realize the history of WP:A and the motivation, possibly temporary, to clarify that each piece of it is a policy, but I still see a potential for confusion. Crum375 03:56, 11 March 2007 (UTC)

Well, so far you're the only one with who's voicing solid objection, maybe we boldly do a trial run here on ATT and see what others think? (→Netscott) 03:59, 11 March 2007 (UTC)

I think it does have significant implications for every policy site-wide. I therefore suggest we discuss it with a wider forum - I have a suspicion that only a few of us are actually looking at this issue right now. Crum375 04:04, 11 March 2007 (UTC)

"Official policy" seal

65 pixels

40 pixels

28 pixels

In SVG format, that scales well and text is readable in small sizes. Hope it works. ≈ jossi ≈ (talk) 02:41, 12 March 2007 (UTC)

Version with text shadow:

At my resolution, the 28-pixel size is unreadable and the 40-pixel size is barely readable. I don't think that it's a good idea to include text in the icon, and I agree with Netscott that the familiar green checkmark should be used. —David Levy 03:12, 12 March 2007 (UTC)

Here is a diff of how the {{policy}} template looks with the seal. ≈ jossi ≈ (talk) 05:04, 12 March 2007 (UTC)

I find this "seal" pretty loathesome. I'm not sure why. It looks kind of like some sort of "Kid Safe!" seal on bin of PlayDoh or something, and I think it sends the wrong underlying message whatever it would look like - that some far-away body of comptrollers in an ivory tower has declared something inviolate and sealed it forever and with regal authority (look into the origin of "seal" in this context and you'll see why that implication is present). WP policy doesn't actually work that way. — SMcCandlish [talk] [contrib] ツ 23:49, 18 March 2007 (UTC)

I agree. The many-pointed edge makes it look like something you would see in a print ad containing lots of exclamation points. Thumbs-down; sorry. —mjb 04:23, 20 March 2007 (UTC)

15px green checkmark, text link

Here's my proposed format:

==No original research==

===What is original research?===

Material counts as original research if it:

• introduces a theory, method of solution, or any other original idea;
• defines or introduces new terms (neologisms), or provides new definitions of existing terms;
• introduces an argument without citing a reliable source who has made that argument in relation to the topic of the article; or
• introduces an analysis, synthesis, explanation, or interpretation of published facts, opinions, or arguments without attributing that analysis, synthesis, explanation, or interpretation to a reliable source who has published the material in relation to the topic of the article.

—David Levy 03:12, 12 March 2007 (UTC)

The green check mark may be familiar to long standing Wikipedians, and graphically only says "check", not "official seal". Here is a version of the seal that has not text, that could be used in the sho7rcut boxes:

≈ jossi ≈ (talk) 03:24, 12 March 2007 (UTC)

To me, a "seal" is just another arbitrary symbol. I don't see how it better conveys the "policy" concept, and I see no valid reason to abandon the setup (green checkmark for policy, blue checkmark for guideline, red X for rejected/historical page) that's been in use since December 2005 and copied by various Wikimedia sites in different languages. —David Levy 03:36, 12 March 2007 (UTC)

A "seal" as a "seal of approval" could be a more effective way to denote official policy, while a check mark says very little. In any case, I was just trying to improve on the long-standing version, and if it can be improved then why not? (I can easily create different color versions for guidelines and for policies) Let's see what other editors think. (≈ jossi ≈ (talk) 03:39, 12 March 2007 (UTC)

To me, the checkmark/x mark symbolism is quite clear, and it would be confusing to change a setup that has become recognizable throughout this site and numerous others in various languages. That consistency is highly beneficial. —David Levy 03:50, 12 March 2007 (UTC)

Sure ... to long standing Wikipedians that already know the policies by heart... But that only represents a tiny percentage of users. ≈ jossi ≈ (talk) 03:53, 12 March 2007 (UTC)

Save for the fact that the green check is already established a seal does make sense in that it is essentially a worldwide recognized symbol that I would argue almost all cultures are familiar with. That said the blue symbol here seems a bit kid like (techno anime)... not serious enough... imho... but I'm liking what I'm seeing in terms of a symbol in the shortcuts area... that is essentially what I was thinking of... something unmistakable about a section being policy. What about the symbol but back to gold? (→Netscott) 03:56, 12 March 2007 (UTC)

Is the concept of a seal recognizable to more cultures than that of a checkmark/tick? (That's a serious question. I don't know the answer.)

I agree that this particular seal seems kid-like. (No offense intended. I wouldn't be able to create an SVG anywhere near as good.) It also is too busy, and I don't like the idea of duplicating text from the template (or introducing an element requiring localization before wikis in other languages could even consider adopting the icon). —David Levy 04:05, 12 March 2007 (UTC)

I see from our checkmark article that the symbol is not commonly used in Japan, but the Japanese Wikipedia, Wiktionary, Wikibooks and Wikisource all use this icon in their policy templates. That it symbolizes an error in Finland and Sweden is of greater concern, but there might be countries in which a seal has little or different meaning. (I have no idea.) —David Levy 04:16, 12 March 2007 (UTC)

Anyone with an SVG editor, such as the open source Inkscape, could localize the text. ≈ jossi ≈ (talk) 05:01, 12 March 2007 (UTC)

I realize that, but I prefer the idea of an icon that can simply be used (and I see absolutely no benefit to duplicating text from the actual template). —David Levy 05:17, 12 March 2007 (UTC)

I perceive the symbols themselves as inherently meaningful. Speakers of no fewer than 26 different languages evidently agreed, and I was referring to consistency with these wikis (not with our own). —David Levy 17:51, 12 March 2007 (UTC)

Any ideas for a different symbol? I am sure we can improve on the checkmark. ≈ jossi ≈ (talk) 04:57, 12 March 2007 (UTC)

What about a stamp? "Stamp! Official". ≈ jossi ≈ (talk) 04:58, 12 March 2007 (UTC)

Note that a check mark denotes the just the concept of "yes" or "validated" and it is very much of US-centric use. ≈ jossi ≈ (talk) 05:00, 12 March 2007 (UTC)

Not sure what to tell you Jossi, maybe we should make a post over on WP:VPR? (→Netscott) 05:06, 12 March 2007 (UTC)

Okay, I know that the "US-centric" claim is false; checkmarks (known as "ticks" in British English) are used in many countries. As I noted, Wikimedia wikis in numerous languages use this icon in their policy templates. —David Levy 05:17, 12 March 2007 (UTC)

I'm not sure how it was decided that check marks would the symbols used for guidelines and policies. Obviously the check mark is readily recognized in English speaking countries. I think gold seals (as typically found on legal documents, etc.) are a bit more universal though for denoting a "seal of approval"/authority. (→Netscott) 05:22, 12 March 2007 (UTC)

I don't know, but I do know that checkmarks are far from US-centric (or even Anglocentric). This icon presently appears in the policy templates of Wikimedia wikis spanning no fewer than 26 different languages (and possibly more if some are using a different filename).

Incidentally, there was no formal decision to use the checkmarks here. I simply added them to our templates and people approved (and editors of the other wikis followed suit). —David Levy 05:46/10:11, 12 March 2007 (UTC)

I see... I will work in creating a Gold version for policies (and a copper version for guidelines maybe?), to replace the blue design that may be too cute. ≈ jossi ≈ (talk) 09:57, 12 March 2007 (UTC)

Okay, but the color isn't my primary concern. —David Levy 10:11, 12 March 2007 (UTC)

Gold version of seal ≈ jossi ≈ (talk) 10:36, 12 March 2007 (UTC)

With caption: ≈ jossi ≈ (talk) 15:44, 12 March 2007 (UTC)

Looking much better Jossi imho..much more "official" looking than either the check mark or the blue seal. The only other thought that occurs to me is that green could be an alternate color in line with the current green check. (→Netscott) 16:05, 12 March 2007 (UTC)

I was thinking actually on "gold" for policies, and "silver" for guidelines. Color is culture sensitive, while Gold and Silver are widely accepted and recognized. What do you think? ≈ jossi ≈ (talk) 16:08, 12 March 2007 (UTC)

Diff of how this will look on the {{policy}} template. ≈ jossi ≈ (talk) 16:12, 12 March 2007 (UTC)

Looking good. My only reservation is that the seal is a bit busy with so many points. If it is not too much trouble maybe you could come up with one that has about half as many points? I agree with you about silver and gold (and copper too for that matter) particularly given the significance of such colors (Olympics, etc.). (→Netscott) 16:16, 12 March 2007 (UTC)

I am running out of time now, but will work on a "less points" version later tonight. In the meantime, other editors can comment further. ≈ jossi ≈ (talk) 16:54, 12 March 2007 (UTC)

No offense, but I see no benefit to the use of a bulky icon (with completely redundant text) that increases the template's minimum size for many users (myself included). —David Levy 17:51, 12 March 2007 (UTC)

Even at 62px, the icon's text is extremely difficult to read on my screen. I don't know what settings you're working at, Jossi, but please try to be considerate of those of us with high resolutions. Mine is 1400x1050 on a 15" LCD.

And again, I strongly object to the use of text in this icon. It's completely redundant and creates unnecessary work for wikis written in other languages.

I believe that you're working on a solution to a nonexistent problem (the false premise that a checkmark is "US-centric"). Given that readers of 26 different languages (not counting English) decided to use the green checkmark, its cultural significance clearly isn't anywhere near as narrow as you've claimed. I see no need or benefit to starting over.

Additionally, the widespread misconception that guidelines are "ranked lower" than policies would be reinforced by the use of such colors as gold and silver (or copper, bronze, et cetera). I deliberately selected a pair of colors lacking such a connotation.

Having given this some thought, I also believe that the connotation carried by the seal imagery itself is inappropriate. Another widespread misconception is that our rules are sacrosanct laws that must always be followed to the letter. To me, this new icon seems indicative of such a situation, while the checkmarks convey the reality that these are mostly consensus-based checklists of concepts determined to be appropriate via use and discussion (descriptive, not prescriptive). —David Levy 17:51, 12 March 2007 (UTC)

How does the gold seal look with the blue style text/lettering? - Denny 16:25, 12 March 2007 (UTC)

Do you mean the text centered rather than around? Or just the same typeface with the shadow? ≈ jossi ≈ (talk) 16:52, 12 March 2007 (UTC)

sorry, that wasn't clear. I think the Gold Seal itself is better than the blue. The current gold seal with the darker lettering doesn't work quite well. make that lettering white (and shadowed) but with the same current wrapping/curved appearance could work. What I was thinking really could be nice however could be the gold seal image with the larger, bolder centered text with shadows you have on the latter of the two blue seals. - Denny 19:32, 12 March 2007 (UTC)

Thanks. I will have a couple new iterations later tonight. ≈ jossi ≈ (talk) 21:11, 12 March 2007 (UTC)

New version: ≈ jossi ≈ (talk) 23:21, 12 March 2007 (UTC)

that is nice. That's what I meant (but you went black instead of white text as suggested--I like your idea better). - Denny 23:28, 12 March 2007 (UTC)

Just to be clear, are you reading my criticisms? You aren't replying to me, and now you've added additional English text and made the intended display size even larger.

The bold lettering helps somewhat, but it remains difficult to read at my resolution (even at the enormous 82px image size).

And again, you haven't demonstrated a problem with the status quo. I've debunked your "US-centric" claim, and you seem to be ignoring this. —David Levy 23:37, 12 March 2007 (UTC)

Debunking? Why should I debunk anything, David? Status quo for the sake of status quo, is not a very strong argument, is it? Or is it that there is no room for improving our graphics? ≈ jossi ≈ (talk) 00:04, 13 March 2007 (UTC)

Please re-read my above comments. I didn't say that you should debunk anything. I noted that I debunked your claim that a checkmark is "US-centric." And no, I'm not arguing in favor of status quo for the sake of status quo. I'm pointing out that your justification for this change (the checkmark's "US-centric" nature) has been proven false. Meanwhile, I've raised various specific objections to the new icon that you haven't addressed. —David Levy 00:35, 13 March 2007 (UTC)

Much improved there Jossi, the seal with fewer points is much clearer. While I like the lettering in the image I'm going to have to agree with David Levy. Universality is important. In response to to David's concerns about policies being considered moreso like flexible guidelines I don't see that changing with the seal imagery. The core fundamentals of policies hardly change and I think it is in the interest of the project that policies be taken a bit more seriously particularly if we want the project to be taken more and more seriously. This as we're coming under more and more scrutiny with the Essjay controversy making headlines. (→Netscott) 23:50, 12 March 2007 (UTC)

I am all for our policies to be taken more seriously, but I do not believe that replacing a tick mark with a seal has much to do with that issue. ≈ jossi ≈ (talk) 00:01, 13 March 2007 (UTC)

My concern is not that policies are taken too seriously. On the contrary, it's that guidelines aren't taken seriously enough. It's common for people to argue that they're entitled to disregard any guideline with which they disagree (because "it's only a guideline"). While it's true that guidelines sometimes have more wiggle room than policies do, they generally should be followed unless there's a good reason not to. That's why I selected a pair of colors that doesn't reinforce the notion that guidelines carry no official weight. Gold and silver (or any other "lower" color) would be counterproductive. —David Levy 00:35, 13 March 2007 (UTC)

But guidelines are ... guidelines. ≈ jossi ≈ (talk) 00:57, 13 March 2007 (UTC)

Meaning what? —David Levy 01:02, 13 March 2007 (UTC)

Here is a Diff of how this will look on the {{policy}} template. I will place a notice at the Village Pump to seek comments from the wider community. ≈ jossi ≈ (talk) 00:10, 13 March 2007 (UTC)

• I think the gold seal (as well as the blue one) looks rather silly on a policy page, and at any rate it's not such a good idea to affix any kind of Official Seal Of Authority to our policy pages. WP:NOT a bureaucracy. Keep the checkmark. >Radiant< 12:05, 13 March 2007 (UTC)

Strongly concur with Radiant. — SMcCandlish [talk] [contrib] ツ 00:18, 18 March 2007 (UTC)

Move the dots closer together and make them bigger and it would be a smiley face. Edison 19:48, 13 March 2007 (UTC)

That's because we have such friendly, happy policies here at Wikipedia! Blueboar 19:57, 13 March 2007 (UTC)

Wrong venue

By the way, this random policy page is an unbelievably inappropriate venue for trying to decide how to visually mark official WP policies. — SMcCandlish [talk] [contrib] ツ 23:51, 18 March 2007 (UTC)

Dispute tag

There is an ongoing dispute about whether to change the policy from "verifiability, not truth" to "attributable ... not whether it is true".

I put the following at the top of this page to reflect the fact that there is a dispute and that there is not consensus:

(However, there is a dispute about the change of policy from "verifiability, not truth" to "attributable, ... not whether it is true." See "Role of truth" below. --Coppertwig 16:40, 5 March 2007 (UTC))

Someone has deleted it. Please put it back. I have major objections to the alleged change in policy. My questions in "role of truth" and in "Is this really policy?" have not been answered. There is definitely a dispute here. --Coppertwig 18:22, 10 March 2007 (UTC)

Sorry, but that is not a dispute. The current version that has been worked on for the last fourth months, represents a fair consensus of editors. ≈ jossi ≈ (talk) 18:29, 10 March 2007 (UTC)

As has been repeatedly pointed out to Coppertwig, the two statements mean exactly the same thing. There is nothing to dispute. This is getting tiresome .... see WP:POINT. Blueboar 22:21, 10 March 2007 (UTC)

Agreed. It's becoming disruptive. SlimVirgin ^(talk) 00:45, 11 March 2007 (UTC)

There is no dispute. We changed one word to go from an oxymoronic claim that made some sense at a shallow level but literally meant "truth, not truth" to one that is clearer and says what it actually means. How is this even disputable? --tjstrf talk 00:59, 11 March 2007 (UTC)

"Verifiable" and "true" do not have the same meaning, and "attributable" isn't even the same thing. "Verifiable" means "could be verified". If I cite a scholarly paper, you can go and verify that the paper says what I claim it does. On the other hand, I can say "I had a turkey sandwich for lunch today." It is true, I really did. It's even attributable (to me), I said so. What it is not is verifiable, there is no reasonable way anyone can verify that what I said is true. Seraphimblade ^{Talk to me Please review me!} 01:03, 11 March 2007 (UTC)

The "verifiability, not truth" wording very accurately describes a long standing, and very important, point of policy. ^High_InBC^{(Need help? Ask me)} 01:04, 11 March 2007 (UTC)

To verify is normally to demonstrate the truth of, [1] so Tjstrf is right when he says the previous phrase was almost contradictory, though in the context we were using the word, it was, as HighinBC says, an important part of our core policies. One of the reasons for the move to Attribution was so we could retain what was important but ditch the misuse of the term "verifiability." SlimVirgin ^(talk) 01:09, 11 March 2007 (UTC)

I like the idea of the new wording along the lines of "Attribution to a reliable source, not truth", but this "not whether it is true" seems a bit awkward. ^High_InBC^{(Need help? Ask me)} 01:11, 11 March 2007 (UTC)

If the old and new wording mean exactly the same thing, let's solve this dispute by going back to the old wording, which was shorter. In my opinion they do not mean the same thing. --Coppertwig 01:40, 11 March 2007 (UTC)

They mean the exact same thing, but the new wording fits the name of the new policy better, and "verifiability" was being misunderstood. There is consensus on this. Jayjg ^(talk) 01:50, 11 March 2007 (UTC)

There is not a consensus on this so-called policy. Saying that there is does not make it so. --Henrygb 04:33, 11 March 2007 (UTC)

Neither does saying that there isn't make that so. Hundreds, if not thousands of editors have been through here and seem happy with the change. It has been discussed extensively. A small minority are unhappy with it. This doesn't mean there isn't Wikipedia:Consensus. The policy tags have been on the page for some time now. JulesH 08:04, 11 March 2007 (UTC)

The present wording is far more useful than the previous one. I have come across editors telling me that they can verify a subject's notability because they have met them (cue messy discussion of what is meant by "verifiability"). qp10qp 14:02, 11 March 2007 (UTC)

The present wording takes the amateur's side at Wikipedia:Astronomer vs Amateur. Does the so-called consensus? --Henrygb 22:22, 11 March 2007 (UTC)

Only if the amateur can cite the suggestion that the moon is made of cheese to a reliable source. I.e., no it doesn't. JulesH 22:42, 11 March 2007 (UTC)

This is a herring. Made of Double Gloucester. qp10qp 22:58, 11 March 2007 (UTC)

NASA is a reliable primary source and Google's use of NASA imagery is therefore a reliable secondary source. Go to http://moon.google.com/ then put an Apollo landing site in the centre of the screen and zoom in as far as possible. You may have to adjust the colour settings on your screen. So, as I said above, this page supports the amateur. --Henrygb 00:11, 12 March 2007 (UTC)

What you're suggesting is taking your own interpretation of a reliable source that isn't explicitly backed up by the words of the source. It may look like cheese, but it requires an expert to interpret it and publish explicit results in a reliable source before we could cite it. So, as I say, no it doesn't support the amateur. JulesH 18:52, 12 March 2007 (UTC)

All your arguments are those of the amateur not the astronomer. The moon is not made of cheese, you know it, I know it, and basing everything purely on sources is not how encyclopedias are written.--Henrygb 21:08, 12 March 2007 (UTC)

(unindent) I don't understand your point. Nobody has pointed to any reliable sources that say the moon is made of cheese. There are plenty of reliable sources that document the actual composition of the moon. There is no difficulty writing an encyclopedia article on this subject based on reliable sources. So what, precisely, is the issue here? JulesH 09:36, 13 March 2007 (UTC)

Synthesis section

The plagiarism example on synthesis remains very difficult to follow. I've mentioned this before, but giving more detail:

• The whole background to the case is complicated and not explained. What does "copying references" mean? The reader is left guessing as to what the case is about.
• The claim that "The whole point of this paragraph is the conclusion that, given the Chicago Manual of Style's definition of plagiarism, Jones did not commit it." does not follow from the text quoted in the example. The paragraph cites a definition of plagiarism, and makes no comment on whether or not Jones committed it or not.
• The example is supposed to be illustrating synthesis, i.e. putting together two sourced ideas A and B to come up with an unsourced conclusion, C. None of the conclusions in this passage appear to rely on more than one source. In particular:
  • The conclusion, "If Jones's claim that he always consulted the original sources is false, this would be contrary to the practice recommended in the Chicago Manual of Style..." does appears to rely on one source only - the Chicago Manual of Style. It is not at all obvious how this is meant to be illustrating synthesis.
  • Similarly, the conclusion "The Chicago Manual of Style does not call violating this rule "plagiarism."" appears to rely on one source only - again the CMS. So this is not synthesis either.
  • The third line "instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them.". Again, this is based on ideas from the CMS only - there is no hint of a synthesis with other ideas or other sources.

From my understanding of the original case, I think that the real dispute was over the interpretation of the CMS (or rather, the Harvard student writing manual, which was the source actually referenced). Smith interpreted it one way, Jones the other. Because Wikipedia does not take one interpretation of the "truth", each interpretation needs "Smith says..." or "Jones says..." before it, and an appropriate reference. This is a very straightforward and well accepted application of the attribution policy, and has nothing to do with synthesis. I propose that the example is removed, and preferably replaced with an example that is easier to follow and actually addresses synthesis. Enchanter 23:01, 11 March 2007 (UTC)

See also discussion at Wikipedia talk:Attribution/Archive 12#Proposed new example of synthesis. I agree with you that a simpler example would be better (or a simpler example, plus a link to this complex one elsewhere). I think, though, that one of the points being made in this example is that Wikipedians should not put two sentences next to each other to make a point if the point is Original Research. Maybe in the original, Smith and Jones didn't refer to the Chicago Manual of Style (CMS), but argued using other definitions. By putting the mention of the CMS in the same paragraph, the Wikipedian editor made it look highly relevant -- practically almost drew a conclusion for the reader. That's what's being deprecated. Maybe if the text introducing this paragraph were better written it would go over better. It could prepare the reader for what's coming. It could say "Note how, by putting two sentences together in the following paragraph, the writer strongly implies something." Instead of saying "now comes the synthesis" when actually there's no synthesis in the first sentence of what comes next -- or in fact anywhere within the paragraph. The synthesis to me comes at the very end of the paragraph, when the reader has taken in all the information and suddenly jumps to a conclusion. There are also other complexities which are not needed here. The simplest example that demonstrates synthesis should suffice. I give some hypothetical examples in the discussion at the link I give above.

Here's another example: Suppose one Source A says "In this town, men often walk around shirtless on the street in the summertime, but women always wear tops." Suppose source B says, "Jane, a resident of this town, was interviewed on the street in this town during a march of the Topless Freedom Society." A Wikipedia article should not say, based only on these two sources, "Jane was wearing a top during the interview." A similar example might be constructed using a Barefoot Freedom Society. However, this example doesn't illustrate juxtaposing two sentences to make a point.

Here's yet another example: Suppose source A says "Any liquid colder than 0 degrees C can be used for this purpose." and source B reports, in a different context, that liquid nitrogen is normally stored at -196 degrees C. Based only on these sources, a Wikipedia article should not juxtapose sentences to say: "Any liquid colder than 0 degrees C can be used for this purpose. Liquid nitrogen is normally at -196 degrees C." because this implies that liquid nitrogen would be suitable for the purpose. --Coppertwig 01:41, 12 March 2007 (UTC)

For the examples, it would be nice to find something that looks more like it could actually be from an encyclopedia article. Coming up with good examples can be surprisingly hard!

Your interpretation of the case as being that the CMS is being introduced when it is not relevant makes sense, but does not appear to be the issue in the real life case - both Smith and Jones had discussed the definitions of plagiarism in the CMS. And in any case, the example is meant to be illustrating synthesis, not the introduction of irrelevant sources. Enchanter 09:09, 12 March 2007 (UTC)

I've now removed the plagiarism example, for the reasons given above. Finding an alternative example would be possible, but maybe isn't that important - none of the other sections of the policy are illustrated with examples, and the explanation of synthesis is reasonably clear without an example. Enchanter 21:42, 14 March 2007 (UTC)

Don't remove material, please, and especially not long-standing stuff like this. You'd need a clear consensus to remove it. It serves to illustrate a point that a lot of editors just don't get at first, and it's a real example, which is perhaps why it seems confusing. SlimVirgin ^(talk)

Could you address the concerns raised above? I understand that, as the person who originally inserted that example, that you find it easy to follow. However, I personally find it very hard to follow, and similar concerns have been raised repeatedly by other users. Firstly, the example is hard to follow for the casual reader, because it is based on a complicated case. And secondly, if you follow the logic carefully, it still does not make sense. The goal of a policy page should be to explain policy clearly and simply, so that readers understand the policy. I don't think the current example achieves this at all

Similarly, the fact that it is based on a real case may be of interest to you, but for new readers of the policy page who are here to find out about Wikipedia policy it does not matter at all whether the example is real or made up. What matters is whether it helps them to understand the policy, and it's here that the example fails. Enchanter 22:03, 14 March 2007 (UTC)

To reiterate, in my view the example currently does not make sense - it's ambiguous and hard to follow for the casual reader, and does not logically hang together if you follow it through carefully. I again propose that it is removed - we shouldn't keep a bad example in the policy page merely because it has been there a long time. Does anyone have any specific comments or objections? Enchanter 23:29, 15 March 2007 (UTC)

I've again removed the example, as none of the problems with it have been addressed. Enchanter 05:23, 17 March 2007 (UTC)

Slimvirgin, I see you have changed this back. Please could you respond specifically to my concerns above, to explain whether and why you think my understanding is incorrect. I understand that you are reverting to text that you yourself wrote, so I think it's reasonable to ask you for an explanation of what it is trying to say.

Just to be clear, I'm not proposing any changes to the actual policy. My concern is that the example in its current form is ambiguous and confusing, does not stand up logically to scrutiny, and does not explain the policy clearly. Enchanter 06:13, 17 March 2007 (UTC)

Strongly concur with Enchanter that the example is pointlessly longwinded and decidedly nonhelpful for the reasons in his/her analysis. If editors are going to editwar to preserve it just because they want an example there then they need to markedly improve the example or (more likely) replace it with one that is actually more useful. — SMcCandlish [talk] [contrib] ツ 16:27, 17 March 2007 (UTC)

The example is bollixed in another way: With a minor edit it would not be OR at all. Simply noting the cited facts from which the editor drew the "forbidden" conclusion and allowing the reader to come to that rather obvious conclusion on their own is actually good editing. All it takes is a minor rewording to get rid of constructions like "if" and "since". Meanwhile, this example appearing in the policy misleads editors into thinking they can't do something that simple to repair a passage like this! Yeesh. — SMcCandlish [talk] [contrib] ツ 17:05, 17 March 2007 (UTC)

Again, I propose that the example is removed for the reasons above. Before I make the change, does anyone have any further comments, objections, or alternative suggestions? Enchanter 19:44, 18 March 2007 (UTC)

I think it would be ideal if a better, replacement example were installed, but I do not think at all that the lack of one should hold up removing one this misleading, complicated and confusing. With regard to some of the earlier discussion, I think that the eventual replacement would be better as a made-up one that only addresses the issue at hand. Searching for another real-world example would likely introduce additional consideration and confusions the way this one did. — SMcCandlish [talk] [contrib] ツ 20:43, 18 March 2007 (UTC)

This will need to be replaced with another example if removed. ^High_InBC^{(Need help? Ask me)} 20:54, 18 March 2007 (UTC)

I agree that one or more examples would be desirable, and would support adding one or more examples to the FAQ, where there would be more room to expand on the examples. I suggest adding one along the lines of the "violent crime" example discussed here, although many other examples could be thought up which could do the job. Enchanter 21:39, 18 March 2007 (UTC)

Works for me, though I reiterate that it is far worse to have a misleading and confusing example than to have no example. This is much like the general WP principal that it is better to remove false information from an article, without replacing it with the correct information, than to leave the false information in place. — SMcCandlish [talk] [contrib] ツ 23:39, 18 March 2007 (UTC)

I've now removed the example again for the reasons given above, and proposed an alternative example to be added to the FAQ at Wikipedia talk:Attribution/FAQ.Enchanter 22:08, 19 March 2007 (UTC)

It's a brilliant example which helps stifle the worst kinds of synthesis. It's clear and concise; there is nothing misleading or confusing about it. Please stop trying to weaken the policy by removing longstanding examples. Thanks. Jayjg ^(talk) 00:13, 20 March 2007 (UTC)

I agree that it's a very good example, which newcomers really need to digest. The only thing I don't like about it is the tone of the line:

Here is an example from a Wikipedia article, with the names changed. The article was about Jones:

I find it a bit too informal and chatty. It's almost like using second person pronouns in an academic essay: it sticks out a bit. I'd prefer something like: "The following example from a Wikipedia article shows that . . . ." well, whatever it shows! But I don't think it would be a good idea to remove the example, as it's a particularly helpful one. ElinorD (talk) 00:49, 20 March 2007 (UTC)

ElinorD, Jayjg, please could you explain in more detail why you think this is a good example? I started this thread with some specific comments about aspects of the example which I think are confusing, ambiguous, and do not follow logically. Please could you explain whether you disagree with these specific points, and if so, why. Enchanter 01:06, 20 March 2007 (UTC)

Slimvirgin, I again ask, please could you give your specific response to the issues raised above? You have reverted the article after I have specifically asked on the talk page whether anyone had any objections to the change, and given people several days to respond. I asked you specifically, twice, to explain your reasons for supporting the policy as it is, and I also left a message on your talk page two days ago. You haven't responded to any of that, and have reverted again without discussion.

I noticed that too. Randomly reverting things without justification seems to be a pattern around here. The flaws in the example have been identified, have not be convincingly argued away, and are agreed up on as flaws by more than one editor. SV seems to demand consensus to delete, but that's backwards; there's no longer any consensus to include and no one, SV or otherwise, has justified the inclusion of the example others think is fatally flawed. Not to mention that we have a FAQ for a reason, and examples, especially long ones, belong in the FAQ not the policy, meanwhile what example to include with regard to this particular part of the policy in the FAQ is in fact a matter of current discussion. I restored the removal of the confusing example. — SMcCandlish [talk] [contrib] ツ 01:21, 20 March 2007 (UTC)

Predictably, someone reverted my revert with an edit summary of "this has been part of the NOR policy for a long time..." as if that means anything. This is not the NOR policy, it is the ATT policy. Consensus can change. Wikipedia is not a legal precedent system. More to the point though, why on earth would anyone want to keep an example that has been flagged by multiple editors as worse than useless, and editwar to keep it in, especially when the policy page is no longer the place for explanatory examples anymore in the first place?! The FAQ exists for a reason. This is turning into "I'm going to revert and preserve the status quo just because I like reverting and I don't like change" game. — SMcCandlish [talk] [contrib] ツ 02:40, 20 March 2007 (UTC)

You have asked that I ensure there is a consensus. A fundamental way of establishing consensus is to ask if there are any objections before making a change, so that people can give their specific thoughts and we can work together to an outcome that satisfies everyone. If you fail to raise these objections at this appropriate time, and revert later, the consensus process breaks down. We need to address concerns carefully, with reasoned arguments and discussion, to make the policy pages as good as they can be. At the moment you are not engaging in the consensus building process, and simply reverting a change that you don't like without explaining why. Enchanter 01:06, 20 March 2007 (UTC)

The plagiarism synthesis example has been in the NOR policy, now incorporated into ATT, for a long time. It is an important part of the NOR aspect of ATT, as it helps elucidate a concept that unfortunately many editors routinely miss. It belongs in the main text, not in the FAQ, as it is critical to the understanding of the 'no synthesis' principle. Given its importance, the only way to remove this example or even replace it with another, would be to attain a clear and wide consensus, that should include the main contributors to this policy page. Crum375 02:50, 20 March 2007 (UTC)

WP:NPOV

I understand you are one, if not the, originators of this essay. Your input, as to, if the scientific POV should be privileged or not would be greatly appriciated. Thx, Fossa_?! 23:45, 11 March 2007 (UTC)

Role of truth

This is a continuation of an ongoing discussion, the earlier part of which has been archived at Wikipedia talk:Attribution/Archive 12#Role of truth.

As I said before, although I'm OK with the change from "verifiability" to "attributable", I worry about the change to "not whether it is true" without a word such as "verifiability" to counterbalance it. I think it would be closer to the original meaning of "verifiability, not truth" to simply delete "not whether it is true" or to insert "merely".

I gave one example earlier of a situation where I think the new policy would produce undesirable results; I'm waiting for those who favour the words "not whether it is true" to comment on that situation.

Here's another hypothetical situation where the "not whether it is true" wording wouldn't be good: Suppose several publications have been reviewed by publishing houses in the normal way, but that the authors and reviewers all come from the same religion, ethnic group or political group, and three publications, of the type normally considered reliable by Wikipedia standards, all state that at a certain time and place a UFO landed and aliens came out and shook hands with people. Suppose no other publication mentioned the incident -- it isn't considered important by other publications. Now suppose someone insists that a Wikipedia article state, "A UFO landed and ...". And suppose other Wikipedians argue that it either shouldn't mention it at all, or should use a prose attribution. The first Wikipedian could use this policy, with its "not whether it is true" wording, to support keeping the statement in as an assertion by a Wikipedia article, without a prose attribution. The user could challenge others to find a publication that denies it, saying that only if such a publication is found must a prose attribution be added. ("prose attribution" means inserting something like "Fun and Games Magazine states that ...")

Here's another example. Suppose three sources report a person's birth date as 1901 and one source reports it as 1801, and many sources report the person doing things in the 1900's and no sources report the person doing anything in the 1800's. Someone could use the "not whether it is true" provision to insist that the 1801 date be mentioned with approximately equal weight (or precisely one-third the weight) as the 1901 birth date. Thus the article would not say "J. Smith was born in 1901," but would say "According to several sources J. Smith was born in 1901, while another source reports the birth date as 1801." People could argue that it's ridiculous to include the 1801 date, but they could be overruled by this policy which states that whether something is true is not the issue.

Here's another example. Now, normally, when someone says something, it is assumed that the person is asserting it, and if it isn't true and the person knows that, the person can be accused of lying (or fraud, or perjury). But if the person says "What follows is fiction." then that doesn't apply. Well, the wording "not whether it is true" could be taken as a grand "What follows is fiction" applying to all of Wikipedia. Someone might take up the hobby of scouring books for typos and misstatements and whenever they find one, inserting it into Wikipedia. The "not whether it is true" wording can be understood to be encouraging this behaviour. If the pattern of the person's edits is noticed and the person is accused of disruptive behaviour and of wilfully writing false material into Wikipedia, the person could point to this policy and argue that the person has a right to continue doing the same sorts of edits on the grounds that it is not the purpose of Wikipedia to present true information, and that others should leave the attributed misstatements in the articles. The person could claim that the person is helping to write the encyclopedia according to its stated purpose.

In reply to SlimVirgin: I don't understand the phrase "do truth". Perhaps you would be willing to explain what you mean in different terms. Re your second comment: Either the new wording means the same as the old or it does not; you can't have it both ways.

In reply to BlueBoar: Of course Wikipedia does not care: Wikipedia is a thing, not a sentient being. It's not capable of caring about anything. But Wikipedians do care about the truth, often passionately enough to engage in edit wars about it. I suppose by "accuracy" you mean accurate representation of what is in the sources as opposed to conformity to reality. I think the dictionary definition of "accuracy" is conformity to reality, so you might do well to explain what you mean when you use the word that way -- I'm still not quite sure that's what you mean. As I see it, we should care about conformity of assertions in Wikipedia to reality as high a percentage of the time as feasible, and requiring attribution is simply a method of achieving both high conformity to reality and less time spent on edit wars. What do you think the purpose is of trying to write an encyclopedia that conforms to other publications and not necessarily to reality? The purpose of writing an encyclopedia that conforms to reality as much as possible is to make a useful reference that people can get useful information from.

In reply to Crum375: I don't think there's a platonic version of the policy somewhere of which the wording on the policy page is merely an approximation. I think the wording on the policy page is the policy itself. If different people had different interpretations of "verifiability", then it's not the place of one Wikipedian to say that one of the interpretations is right and another wrong, (or that one is "true" and one is "false") especially when large numbers of Wikipedians had the same interpretation. If the words meant one thing to large numbers of Wikipedians, we can't assume there was a consensus supporting a different meaning.

In reply to Jakew: Why do you say we're not in a position to determine whether a statement is true or not? Wikipedians come from many walks of life; probably many are the authors of material used as reliable sources by Wikipedia. Do you mean we sometimes are not in that position, or that we can never determine whether a statement is true? Can we determine whether the following are true?

"We're not in a position to determine whether a statement is true or not."

"This book contains this statement on page 53."

"There was consensus in favour of the new policy."

I think we often are able to determine whether a statement is true, at least with a high degree of certainty. The problem is that sometimes people disagree about what is or is not true, and requiring attributions is a good way to resolve those disagreements, since attributions are often more objective than other things people might bring into an argument (e.g. personal experience); people tend to agree more about what the sources say than about whether what the sources say is true. So that's fine to require attributions. But it doesn't follow from that that we have to claim we never know what's true or that we have to change the whole purpose of the encyclopedia.

In reply to Mackan79: Excellent point. Without the word "verifiability", there's no more need to say "not truth" than to say "not whether it is relevant, interesting, important" etc.

Would those who support the wording "not whether it is true" please provide an example of a real or hypothetical situation (such as a content dispute) where you believe that wording would be helpful? Please explain why those particular words would be helpful, not just the change from "verifiability" to "attributable".

I note that no one has specifically addressed some of the specific suggestions I made for addressing my concern. For example, no one has objected to my suggestion to insert "Not all attributable statements are worthy of inclusion." If there are no objections, I will soon edit this sentence into the policy. --Coppertwig 00:03, 12 March 2007 (UTC)

Coppertwig, as for what I mean by "accuracy" ... I mean that we should report what the sources say accurately. This means we should examine the entire source to be sure that it does indeed say what we claim it says. We should also examine a source to make sure that it states the facts accurately. If it misquotes someone, or takes a quote out of context, then we have good grounds for determining that the source is unreliable. These are issues that must be hashed out on the talk page of each article.

Lets take your 1901 vs 1801 example... if there is honest debate as to these dates, then yes, we do need to report that one source says the person was born in 1801 while two sources state he was born in 1901. In reality, if there was a range of 100 years between the reported date of someone's birth, the editors would know that there was probably something wrong with one of these dates and hash it out on the talk page of the article... some common sense would probably come into play. For example, if the source that says the person was born in 1801 later states that he died in 1978, at the age of 77, then it is fair to assume that the 1801 date was a typo. There would quickly be a consensus that the article should disregard the first date as unreliable. But, let's take a more realistic situation: say two sources say a medieval King was born in 1355, but another source says he was born in 1353... we should indeed report the fact that the date of his birth is debated, and give both dates. Wouldn't you agree?

As for your UFO story... (assuming it met the criteria of WP:FRINGE which usually covers such claims) if the landing of a UFO is reported in a reliable source then, yes, according to this Policy we would indeed have to say "A UFO landed..." In reality, since UFO sightings are rarely if ever reported in reliable sources (they tend to be the stuff of the marginal press) it would be proper to attribute the statement as "According to a report in UFO Magazine a UFO landed..." This is a statement that accurately relfects the facts and fits with the parameters of this Policy. It is not up to us to determine if a UFO actually did land... only that this was reported.

As for "What follows is fiction"... if the statement is relevant to the article, then yes, we should report that the author made the statement and note that he stated that is was fiction. Blueboar 13:12, 12 March 2007 (UTC)

Thank you so much for replying. I agree completely about how we ought to be reporting birth dates and hashing out on the talk page whether a birth date given in a source seems to be a mere typo. In other words, we need to concern ourselves with whether the material presented in the sources seems to be plausible and to accurately conform to reality as far as we can reasonably tell. This is why I would like to delete "not whether it is true" from the Policy. It seems to me that the policy, with the "not whether it is true" wording, could be used to avoid any discussion of whether 1801 is a typo and to insist that 1801 be mentioned in the article regardless of whether it could possibly be true.

I agree that the policy as currently written could be interpreted as requiring Wikipedia articles to state things like "A UFO landed ..." if the sources, of a type normally considered reliable, say so. I see that as an undesirable aspect of the policy as currently written. What do you think? Do you think it's productive for Wikipedia to report that sort of thing, and if so, why?

I think you missed my point about "what follows is fiction". I was not talking about an author of a source saying that. My point is that the "not whether it is true" wording in this policy in effect states that all of Wikipedia can be taken as being fiction, and in so stating, it frees Wikipedian editors to insert known falsehoods. I see this as a major flaw in the policy as currently written.

What if several books printed in a certain country state that the leader of that country physically shook hands with God, and those books are the type of source normally considered reliable sources by Wikipedia and have no other surprising information, and there are no published sources stating that the leader did not shake hands with God. Should Wikipedia report "The leader of this country has physically shaken hands with God."? I don't think so. At most it should be, "Several books state that ...". In other words, Wikipedians do need to concern themselves with whether the statements made in Wikipedia articles seem to be true as far as the Wikipedians can tell using their common sense. --Coppertwig 22:20, 12 March 2007 (UTC)

I would think that a claim that the leader shook hands with God would be a significant part in an article on him. We should indeed report that these sources say he physically met with God... Chances are, if such a claim were included in three reliable sources, then the leader is tied into his countries religious practices and belief due to this claim. Wouldn't that be worth mentioning? Have you never tried to edit an article on a religious topic? See: Muhammad as just one example. You might not agree that what the islamic texts say is "true"... but could you really see not mentioning what they say in the article? In fact, going back to your example, if you didn't include the claim that the leader physically met with God, you would have a hoast of editors from his country, his followers, yelling at you for violating NPOV. Blueboar 00:30, 13 March 2007 (UTC)

I see no objections here to the addition of this sentence in the 12 days since I proposed it: "Not all attributable statements are worthy of inclusion." This is one of the suggestions I made to address the concern being discussed in this section. Although there is no objection expressed here, someone has reverted the addition of the sentence. Please don't revert edits without discussing on the talk page. The editor who did this revert was specifically invited by me to participate in discussion in this section and had expressed no objection to the sentence, nor suggested any alternative way of addressing the concern. Discussion should take place here on the talk page, not via edit summaries. Would anyone who objects to the sentence please express your objection here, and discuss how to resolve the dispute. You have already has 12 days to discuss it. --Coppertwig 21:02, 15 March 2007 (UTC)

If people don't respond, it probably means they don't agree. From now on, please take that to be the case with me. You can't change the policy without clear consensus (actively expressed) from the editors who are involved with it. SlimVirgin ^(talk) 21:05, 15 March 2007 (UTC)

It hasn't been my observation that no one makes changes to this policy without "actively expressed" agreement, nor that the "default" is disagreement. In any case, since I share Coppertwig's and (Beit's) concern and his suggested solution, may we please hear any objections thereto and alternative ways of addressing the concern expressed? Lethiere 22:25, 15 March 2007 (UTC)

If Coppertwig had not proposed including this sentence, I would have done that myself. I've seen the argument "this is sourced to a reliable source; you can never ever remove it" so often that it's getting sickening. It is very important that the policy gives this obvious editorial advice clearly and unambiguously. Beit Or 21:12, 15 March 2007 (UTC)

(unindenting) Unfortunately, that sentence doesn't say anything. We can similarly add "Not all sentences beginning with the letter A are worthy of inclusion; though some are." It's clear that not everything from a reliable source must be included, otherwise we'd have to copy whole books into our articles, but this isn't the place to write that. --AnonEMouse ^(squeak) 21:20, 15 March 2007 (UTC)

Unfortunately. what is clear to you is not clear to many Wikipedia editors; at least, this has been my experience. This sentence is powerful reponse to "it's sourced, don't remove it" argument. People will have to find better arguments for includion of material than the fact that it is attributed to a reliable source. Beit Or 21:26, 15 March 2007 (UTC)

It is important that anyone who disagrees with the proposed change be willing to discuss their objections here. SlimVirgin and Crum375 have reverted the proposed sentence, against the consensus on this talk page, but have not articulated here any reason for preferring the version without that sentence. We still haven't heard here, if I remember right, from anyone who supports the new, unqualified "not whether it is true" wording an example of a situation in which they think that wording would have a beneficial effect in comparison to not having those words.

Please do not revert the new sentence without expressing clearly here the reasons for opposing it. We're still in the process of trying to find a wording that consensus can be built around after the new, unqualified "not whether it is true" wording was proposed. We need everyone involved to state their opinion clearly so that we can search for a wording that everyone is reasonably satisfied with. Reasons have been stated for including the sentence "Not everything which is attributable is worhty of inclusion." No reasons have been stated here for not including this sentence.

I'm sorry, but I'm not willing to interpret silence as objection. If people have objections, they need to communicate not only the fact that they have an objection, but also the reasoning behind the objection, so that a discussion can continue from their in search of a wording that everyone can live with. I've noticed that some changes have been made recently to the policy without my having commented in advance, so I don't think taking silence as objection is the norm. We also need to consider the silence of those who used WP:Verifiability for many months, without seeing any merge tag on it, only a notice that it was policy. If such people are now reverting the new wording in various ways it should not be surprising. --Coppertwig 23:29, 16 March 2007 (UTC)

In reply to BlueBoar: I was imagining that the statements that the leader physically shook hands with God would be a minor part of each of the books, somewhat more prominent, but not much more prominent, than inserting "may he live forever" after someone's name. In any case, a clear distinction needs to be made between Wikipedia mentioning something with a prose attribution, and a Wikipedia article stating that something occurrred; that is, between "Three books say that the leader shook hands with God" and "The leader shook hands with God." My question to you is: Should the Wikipedia article state "The leader shook hands with God."?

Here's another example. I believe in a recent year I heard a (normally) reliable news service report that a strange flying object had been seen in the Northern sky. Updates were given as the object reportedly got closer. It was reported as being pulled by flying reindeer, and with a man in a red suit at the back. The date was Dec. 24. If several reliable news services make similar reports, and if no news service reports "There is actually no flying object pulled by reindeer as reported by other news services tonight", then should a Wikipedia article state "A flying object pulled by flying reindeer was seen in the sky."? (Note, I am asking whether a Wikipedian article should state this, not merely whether it should mention it.) Or, should Wikipedian editors pay some attention to whether the material being reported by the news service actually seems to be true? --Coppertwig 23:44, 16 March 2007 (UTC)

It doesn't look to me as if there was a consensus.

CMummert said on Feb. 23, "ATT has been tagged as policy for 8 days, and as far as I can tell there are no outstanding objections. Perhaps it is time to implement the redirects? ... CMummert · talk 17:45, 23 February 2007 (UTC)". But in the 8 days previous to this, six different users made the following six comments (extracts of the comments are shown, and in most cases diff links), and with the exception of possibly DCB4W, I see nothing to indicate that they had retracted their objections:

And perhaps if WP:V redirects to it, mention the word "verifiability" at least once in the policy? Gracenotes^T § 05:48, 16 February 2007 (UTC) (At Village Pump. See [[Wikipedia talk:Attribution/Archive 12#Case in point: WP:ATT]].)

Are you really suggesting that an editor who knowingly includes cited, but false, information, is helping write the encyclopedia? ... DCB4W 16:20, 17 February 2007 (UTC) [2]

(Incidentally, I'm still waiting to hear an answer to this from those who support the new wording. Coppertwig)

I think it might be best to just keep the name it "Verifiability". —Centrx→talk • 19:33, 18 February 2007 (UTC) [3]

"Strong objection" A435(m) 22:29, 22 February 2007 (UTC) [4]

[5] "It seems to me that an overly strict reading "The threshold for inclusion [is] not whether it is true" will open the door for unscrupulous editors to convey attributable information that they know to be factually inaccurate as simple statements of fact. I cannot help but think that this would compromise the reputation of our project. CJCurrie 03:27, 23 February 2007 (UTC)"

"Having "attributablity-not-truth" embedded into policy gives the upper hand to pushers of fringe theories. Wouldn't it be better to water down the opposition between attributability and truth here with a "not merely" or "not necessarily"? that is, to identify "truth" as a necessary but not sufficient condition? semper fictilis 15:10, 23 February 2007 (UTC)" [6]

--Coppertwig 00:31, 12 March 2007 (UTC)

The whole verifiability/attributability vs. truth things make sense if you distinguish between fact (i.e. undisputed) vs. opinion (disputed), or better yet the varying degrees of disputedness, and point people in the direction of WP:NPOV. The main point is, it is not Wikipedia's place to judge what is true and what is not. Trying to do so can lead to very ugly content disputes. We should document disputes without engaging in them... at least not on the article. Armed Blowfish (talk|mail) 01:00, 12 March 2007 (UTC)

I contend that it often is the place of Wikipedians to judge what is likely or apparently true and what is likely false or contentious. An article in Wikipedia should not say "God really exists." but it can say "Acquinas argued that God really exists." On the other hand, an article in Wikipedia can easily say "P. Jones was born in the year 1901", but it would probably be awkward and inappropriate for the article to say "A recent biography reports P. Jones's year of birth as 1901." In most Wikipedia articles that I've seen, most information is presented in the form of ordinary sentences without prose attribution. To do this requires judgement on the part of Wikipedian editors about the likely truth value of the information.

Wikipedia needn't report whose account is being used for P. Jones' year of birth only if there is no notable dispute about it. If there is a dispute, then Wikipedia would have to report whose version of events was being supplied, just as it has to regarding the existance of God or any other notably disputed fact or claim. Only a small percentage of the world's claims are ever disputed, particularly when it comes to mundane details, so most of the time there's no need to worry about mentioning the attribution in the article's text (as distinct from its footnotes or sources). --Shirahadasha 16:47, 14 March 2007 (UTC)

I expect uglier content disputes with the "not whether it is true" wording. The reason is that if the policy flies in the face of common sense, some editors will passionately follow the policy while other passionately follow common sense, leading to edit wars. It's better to have a clear, sensible, easily followed policy such as simply requiring attribution without saying anything about truth. I don't see how that would lead to any more edit wars than any other policy. When Wikipedians disagree about whether something is true or not, they simply put in a prose attribution, making the assertion in the Wikipedia article trivially true (i.e. it's true that the source said that.) --Coppertwig 01:17, 12 March 2007 (UTC)

If Wikipedians try to judge what is true, then quite likely they will argue about whether or not God really exists, and for that matter, which God exists, and what is the correct meaning of God. Not able to agree on the answers to these questions, they may simply revert each other endlessly. If we put aside the question of the truth of the existence of God, we can simply report what our sources say, proportional to the relevance of their opinions, and attributed to their source, and write an article sympathetic to all major points of view. If something is undisputed by all sources that are significant, it can be stated as fact. — Armed Blowfish (talk|mail) 01:28, 12 March 2007 (UTC)

Wikipedians can judge whether some things are true without having to argue about whether God really exists. As soon as they notice that some Wikipedians believe that God exists and some do not, they can start using prose attributions; no argument is necessary. Part of the purpose of requiring material to be "attributable" is to cut down on edit wars.

Do you really think a Wikipedia article should state as fact that a UFO landed and aliens came out and shook hands with people if all 3 sources (sources of the type that are normally considered significant and reliable) that mention this particular incident agree on it? --Coppertwig 01:51, 12 March 2007 (UTC)

I think there was an excellent example above somewhere, it must've got archived but it was a great example. Generally, if something is contentious at all, it's best to state with a prose attribution, if not, we can simply state it. "Abraham Lincoln was the sixteenth President of the United States.^[1]" is fine, but "Scholar X states that Lincoln's suspension of habeas corpus was constitutionally permissible under wartime power based on the Smith v. Doe decision.^[1] However, Scholar Y disagrees, citing the Was Not Either clause, as well as the Whoever vs. Someone case.^[2]." Seraphimblade ^{Talk to me Please review me!} 01:59, 12 March 2007 (UTC)

I agree with Seraphimblade's procedure, which is the one I normally follow. If a remark is bland "Britney Spears was born in Dallas" we can cite it without explicit attribution and merely include a source like "The Life of Britney Spears by xyz" under "References". However if a remark is likely to be seen as contentious "Britney Spears was abducted by aliens three times" then we should attribute it explicitely to a particular source. "Britney Spears was abducted by aliens three times", according to her mother<ref>"Amazing facts", New York Times, Apr 4, 2007</ref>Wjhonson 02:08, 12 March 2007 (UTC)

This really seems like a Neutral Point of View issue to me. Why not simply talk about the distinction between facts and opinions (which are unrelated to truth), with a nice link to that policy, tying other parts of WP:ATT into WP:NPOV if we feel like it? — Armed Blowfish (talk|mail) 02:13, 12 March 2007 (UTC)

I'm not sure NPOV covers it completely. I mean what about the case where an "assertion of fact" is neither positive nor negative? "Britney Spears says banana pie is her favorite pie". Wjhonson 02:16, 12 March 2007 (UTC)

I think it is good to talk about how sources affect neutrality, but I think WP:NPOV has a lot to say about that example. Suppose, for example, Britney Spears said that banana pie was the best kind of pie. Now, on the pie article, for example, suppose they are writing a section about the best and worst kinds of pie. Britney Spears is not an expert on pies: including her opinion on the pie article would be giving her undue weight on the subject of pies. However, Britney Spears is an expert on her own opinions, so we can include her opinion about banana pie on the Britney Spears article. See WP:NPOV#Undue_weight.

However, even in the Britney Spears article, we cannot say that banana pie is the best kind of pie: we must attribute this disputed opinion per WP:NPOV#Attributing_and_substantiating_biased_statements. Also, when we are discussing the best kinds of pie in the pie article, even better than saying "X thinks banana pie is the best kind of pie" is stating why X believes that. For example, is banana pie generally sweeter or smoother than other kinds of pie? See WP:NPOV#Let_the_facts_speak_for_themselves.

I think WP:NPOV covers the topic quite thoroughly. WP:NPOV#A_simple_formulation discusses the difference between facts and opinions: I think WP:ATT can expand on this by showing how to use sources to tell the difference, but should let WP:NPOV tell what to do about it. — Armed Blowfish (talk|mail) 02:36, 12 March 2007 (UTC) Broke this up into paragraphs for easier reading, 03:10, 12 March 2007 (UTC)

(Incidentally, I'm still waiting to hear an answer to this from those who support the new wording. Coppertwig)

The new wording is no different in this regards to the old wording. A reliably-sourced but incorrect statement is verifiable, according to the definition of verifiable that was used in the old wording. JulesH 19:54, 12 March 2007 (UTC)

Yes, absolutely true. There are some problems with WP:NOTTRUTH, such as the incident with Cohen's Persian descent, where a reliable source copied Wikipedia and this same source was used to perpetuate the falsehood. The important fact is that the cited examples all used WP:VERIFY to justify the inclusion of cited falsehoods. The role of truth has not changed. --Merzul 23:15, 12 March 2007 (UTC)

Coppertwig, please add another example, raised on February 22nd, to the six you enumerate above as unresolved objections to implementation of WP:Attribution as policy. To wit: "I still think this policy should include text that instructs editors to use common sense. Reliable sources publish incorrect information, and unreliable ones publish correct information. If there's a good way of telling the difference (which may be specifc to the particular situation), that should be enough for us. JulesH 09:35, 22 February 2007 (UTC)" -- see Wikipedia talk:Attribution/Archive_11#Can information be both verifiable and demonstrably false? I had raised a variation on this complaint here on February 8th and March 1st as the Diligent dissident's scythe. My concerns were dismissed, as were those of the others, (in particular the cogent arguments of Ken Arromdee in the above-cited discussion, and those of Tearlach in Dissident's scythe).

The common thread in the complaints is that one cannot declare truth to be irrelevant to decisions about what factoids stay and go in WP -- on the grounds that people turn to an encyclopedia because they seek truth, and they count on its editors to hold truth as the ultimate standard for entries. Because editors fully realize that "truth" is a slippery slope, what those concerned have called for here is not the blanket right to appeal to truth, but guidelines and examples clarifying when it may be an appropriate consideration (e.g. undersourced but undisputed statements may be deleted on sight under WP:A, whose wording and tenor emphasize this deletability in a way that was not usually applied under the previous policies).

The response given to this point is invariably, "attribution being the closest encylopedic proxy for truth achievable, it is also WP's goal and standard in lieu of truth, therefore: 1. We don't believe WP:Attribution will permit known untruths to enter WP or known truths to be removed, but to the extent it may, this will be trivial and speedily corrected; and 2. Since our intent was to change nothing in previous policy, this new iteration cannot encourage any more negative or unintended results than the predecessor policies -- so if you think it does, you just don't understand what we've done here." Henceforth the objections raised, no matter how elaborated or documented, are deemed answered and rebutted. They haven't been. But until a broader audience, perhaps of slashed-and-burned WP editors of good faith (as john k becomes in this case), finally turn their attention to this new scythe, the WP:A grim reaper won't be restrained. Lethiere 04:25, 13 March 2007 (UTC)

I would point out that my comment there was not an objection to WP:ATT being implemented as a policy, but a suggestion for an improvement to it. The same weakness also applied to the previous version at Wikipedia:Verifiability. I also have not yet heard a response to my answer to these complaints, that all of them applied equally to the previous policies that WP:ATT has replaced. Except User:Henrygb's suggestion that verifiability hasn't had consensus for over a year now, which I'm pretty sure is wrong -- WP:V has been cited by a large number of wikipedia users scattered across hundreds of talk pages, afd debates, and other debates during this period, and nobody has challenged it until now. So: people who oppose this, can you point out a real, practical problem that this policy has but which the previous one didn't. Because nobody has done so yet. JulesH 13:41, 14 March 2007 (UTC)

I agree that WP:V did have consensus. But there were several different reasonable interpretations of what it meant. I mistakenly spent a long time last fall discussing these differing interpretations in the context of WP:CITE. For the time being, in order to have consensus, WP:ATT has to be a "big tent" policy that can accomodate both:

• editors who feel that the ultimate goal is truth, but accept that true but unattributable material can be removed
• editors who feel that the ultimate goal is to have every statement backed up by a reliable source and feel that truth is irrelevant

Of course these are not the only two possible viewpoints. My point is that neither of these two groups would be able to get consensus for rewriting WP:V or WP:ATT to explicitly reflect their interpretation. This is fine, because WP uses consensus to interpret its policies, not a pseudo-legal system, and there is no bar for wikilawyers. CMummert · talk 14:43, 14 March 2007 (UTC)

I don't object to WP:ATT either -- provided that it addresses the concerns repeatedly raised about not granting license 1. to delete content whose accuracy is not seriously disputed, 2. nor to protect sourced content which is demonstrably (but not necessarily attributably) incorrect.

The expected difference in impact between the old policies, as practiced, and the new one, as promulgated, has been stated repeatedly. For example, I pointed out the potential fallout on Feb 8th, "Again, it's acknowledged that these aren't new policies. But the point and effect of refining and re-iterating them is to emphasize their applicability and to legitimate all intolerance of undersourced content". At the "practical" level, the difference may be seen (and was cited upthread) in a recent edit-war that grew out of WP:V being cited -- atypically -- for deletion of truthful content, "I've been here for four years, and I've never run into anyone demanding sources for utterly uncontroversial statements the way that you are. If we are obsessive about this kind of thing, wikipedia will never grow, and it is completely counterproductive to revert unsourced statements of what is essentially common knowledge. And I think my understanding of this is a lot closer to the way wikipedia actually works than yours is, which is based on a completely non-existent perfect world idea of what wikipedia should be. john k 05:39, 11 March 2007 (UTC)"

Obviously, we can't yet cite the impacts of WP:ATT, but both WP:V and WP:ATT authorize anyone to delete unsourced content without regard to whether it is true or has been disputed. But in fact, under WP:V, the tacit implementation of that dictum has usually been to overlook or (increasingly) to tag, rather than to delete, such material. WP:ATT implicitly revokes that tacit understanding by 1. failing to incorporate it explicitly instead of blanket authorization for immediate deletion, and 2. by the apparently growing desire, bordering on impatience, among some editors (evident during this process) to enforce the official policy literally and maximally, henceforth if not retro-actively. The result, I predict -- and keep calling attention to -- is the kind of edit-warring mentioned in the case above, where old-school editors butt heads with those of the new regime.

And that's only the problem to be expected when two reasonable editors conflict. When one of an article's editors is a "diligent dissident", I predict that WP:ATT will be used as a scythe to delete or tag-to-death undersourced but widely accepted content -- because the tacit consensus against challenging truth will have been over-ridden by promotion of WP:ATT. Lethiere 00:27, 15 March 2007 (UTC)

I've repeated the senence encouraging civil behavior in removing stuff to the lead; it was down below. I hope this will help keep a culture of restraint. We are the same people, after all. Septentrionalis PMAnderson 02:57, 15 March 2007 (UTC)

Here's another example similar to the "diligent dissident": Talk:Video_game_crash_of_1983#Serious Problems. This one is interesting because the guy asking for sources actually admitted that a similar request for sources was improper, when it was being used against him in an article he liked. Ken Arromdee 13:40, 15 March 2007 (UTC)

Some areas where I've argued against the "consensus":

• Using the Dershowitz (Smith/Jones) example as an example of NOR
• Including false information (see above)
• Not allowing the popular culture exception
• Not allowing wikis as sources even when the objections (self-published sources, can change at any time, hard to determine who really wrote it) don't apply

Consensus? Not in the least.

By the way, "W:V has consensus, because it's been cited a lot" is not valid. It could very well be that some parts, which are cited, have consensus, and other parts don't. It could also be that people are reluctant to argue against a policy and may not be sophisticated enough to know they should point out the policy has no consensus. Ken Arromdee 14:49, 14 March 2007 (UTC)

Advice requested

For a while, Flatbush Depot and a number of similar articles have had information about what types of buses are located there and what bus routes serve the depot. The only source given is a "fansite", and I do not believe that it is possible to find a better source. If I remove the information, I know that it will be reverted. What should I do? --NE2 03:25, 12 March 2007 (UTC)

I would recommend first trying to find a better source, if you have not done so already. If you can't find one, you can try putting {{Unreliable|article|March 2007}} on the article. If no one adds a reliable source in a reasonable amount of time, which will depend on the traffic the article receives and talk page comments, I would recommend nominating it for deletion or merging, documenting your efforts to make the article well-sourced.

Some editors may disagree with me.

— Armed Blowfish (talk|mail) 03:49, 12 March 2007 (UTC)

I have tried to find a better source and failed. I might create bus depots of the Metropolitan Transportation Authority (New York) bus depots of the New York City Transit Authority and then redirect everything there, but the issue is that it will be me against all the regular editors of these articles. --NE2 03:56, 12 March 2007 (UTC)

AfD will bring it to the attention of a broad cross-section of the Wikipedia community, not just the regular editors of those articles. However, if you try putting {{Unreliable}} on the article for awhile, giving other editors a chance to look for good sources, I would think that the AfD would probably go better. Other editors might look for sources in places that you have not thought of, and in the meantime the template will alert the reader not to trust the article too much. I do not know if you can nominate things for merging on AfD, although merging is the outcome sometimes, but there is also Article RfC. — Armed Blowfish (talk|mail) 04:07, 12 March 2007 (UTC)

I'm pretty sure AFD is for deletion only. I don't think the topics are bad (I added some historical information to Flatbush Depot); I just think some of the information is bad. --NE2 04:16, 12 March 2007 (UTC)

I didn't see this brought up at WP:NYCS. Perhaps ask there? Gimmetrow 04:23, 12 March 2007 (UTC)

And so it begins... any advice? --NE2 02:15, 13 March 2007 (UTC)

Never mind... BWCNY seems to have accepted it. Thanks Edison if it was your revert that did it. --NE2 23:03, 13 March 2007 (UTC)

Comments needed

Sorry for intruding, but we could use some WP:ATT expertise at Talk:Johann_Hari#Request for Comment: Disputed Photo. Briefly summarized, the parties dispute whether an image is usable in the Johann Hari bio article - Hari denies that the picture is him, but some editors believe that the image is unmistakably him, and the image is captioned (on Flickr) as being a photo of Hari taken at an event that he later wrote about. Is the picture subject to WP:ATT, and, if so, does it meet the requirements? The full explanation is here, and comments are welcome here. Thanks, TheronJ 13:32, 12 March 2007 (UTC)

I've left a comment. SlimVirgin ^(talk) 22:47, 12 March 2007 (UTC)

Confirming editors' credentials

I'd like to propose adding this new section to the policy in order to deal with the confirmation-of-credentials issue that is being wrestled with in various places. Any thoughts? SlimVirgin ^(talk) 00:59, 13 March 2007 (UTC)

I have always seen the proposals from this ed. to be eminently sensible, but this one needs rethought. How can it be sourced that I was born in Philadelphia without violating privacy. or that another person is heterosexual? or that a third has worked in the past as a railroad engineer? Or that someone else is a member of the Lutheran Church? All factual questions. For people in article space, there is a bio from which documentation can be taken. But most of us are not of that much importance yet. If someone does want to write an article about one of us, then it has to be sourced. I might claim a degree i music from somewhere, but the documentation would at best prove there is someone with my name who has a degree in music. All this has been discussed. For N individuals, such material can be found without COI. But not the eds. (I do know a few eds. who edit under their real name and have bios in here. Let it apply to them. DGG 01:15, 13 March 2007 (UTC)

How about something to the following effect: "User pages and the claims made on them are not encyclopedia articles and cannot be verified. You should not trust such claims any more than you trust the user making the claims." – Þ  01:29, 13 March 2007 (UTC)

Besides the good points DGG makes, I don't like this for the simple reason that this policy is about our content and not about our editors, and it should stay that way. --Conti|✉ 01:40, 13 March 2007 (UTC)

No, it covers academic qualifications and professional claims only. Please look at it again. SlimVirgin ^(talk) 01:43, 13 March 2007 (UTC)

Hmm, yes, but it still would be about editors and not about content. Also, where do we stop with this? When we demand sources for academic qualifactions, why not demand source for claims like "I work at the University of X" or "I work at Microsoft"? Even if I would agree with the general idea behind this, I'd say that WP:ATT (and WP:BLP, for that matter) is the wrong page for this. --Conti|✉ 01:51, 13 March 2007 (UTC)

There seems to be confusion about what it says. Here it is:

Editors should not make claims about their professional expertise or academic qualifications anywhere on the site, including on their user pages, without supplying reliable, third-party source material in support of it. The source must be independent of the subject and should include the subject's real name. If the material has been published, the publisher must be a reliable, non-vanity press. If not published, the source material must be provided by a trusted source such as a university; for example, a degree certificate from an accredited body uploaded onto the site and displayed on a user subpage would satisfy this provision. Any such claims on user or other pages that are not attributed to a trusted source may be removed by any editor, although editors should use their common sense and be sensitive in their approach when considering whether to do so.

SlimVirgin ^(talk) 01:45, 13 March 2007 (UTC)

SlimVirgin: There is a vigorous debate ongoing at WP:CRED. I would suggest to wait and see how that proposal plays out, before adding something here. ≈ jossi ≈ (talk) 01:51, 13 March 2007 (UTC)

I'm having trouble imagining how the proposal would provide real verification. Can you provide some specific and detailed examples of how this proposal would work in practice? --JWSchmidt 03:19, 13 March 2007 (UTC)

A much easier idea, tell everyone to take unsourced statements on userpages with a grain of salt, but do so on another policy / guideline. -- Ned Scott 04:02, 13 March 2007 (UTC)

I agree that the proposal is problematic for the reasons stated above by others. Slimvirgin when you say "professional expertise" that is a very broad statement. How do I prove that I've been an auto mechanic for five years? How do I prove that I have a business, out-of-my-home, decorating birthday cakes? "Professional expertise" is a bit broad. Also credential images can be easily faked. I got a copy of person X's diploma, put my name on it and voila I'm a Doctor of Divinity.Wjhonson 04:08, 13 March 2007 (UTC)

I see three basic problems with the proposal to offer verification for claims of professional expertise.

1. Some people may put information on their user pages in a way that makes information such as their home address accessible to a wide audience, without thinking through the possible adverse consequences.
2. As pointed out above, most of us can only prove that a person with a certain name has certain credentials; we can't make a verifiable link between the name of the professional and the Wikipedia user name.
3. The practice of providing professional credentials will suggest we are moving to do away with the requirement that articles be attributed to reliable published sources, and that it really isn't so bad to rely on the original research of a Wikipedian with suitable credentials. --Gerry Ashton 05:01, 13 March 2007 (UTC)

Just adding my voice to the opposition here. There are thousands of Wikipedia users who already have listings of academic qualifications on their user pages. See Category:Wikipedians by degree for just some of them. Are we really going to require all of these users to provide evidence of these qualifications, despite the fact that in 99% of cases those qualifications are totally irrelevant, and nobody will treat the user differently because of them? User pages currently enjoy an exemption from this policy, and there are good reasons for this. JulesH 08:07, 13 March 2007 (UTC)

Unless the whole content of this page was to be applied to user pages, I don't think the content of user pages should be covered here. A major policy innovation like this needs its own page. --Zero^talk 09:07, 13 March 2007 (UTC)

An attempt guage community support on this and related proposals is going on at User talk:Jimbo Wales/Credential Verification. Please participate. Thank you. WAS 4.250 11:33, 13 March 2007 (UTC)

Leaving asside the issue of whether verification of claimed credentials is something we want to do ... We have long held that the policies that were merged into this one only aply to articles, and not to user pages, draft articles attached to a user's page, or an article's talk page. If verification of credentials is something we as a community want, it should be made a policy on it's own. Keep this one purely on articles. Blueboar 12:34, 13 March 2007 (UTC)

I think, per Jossi, that WP:CRED is the place to have this discussion. I also don't know how we'd enforce the "should not". Will other editors have a right to remove such info from user pages? That would create a terrible situation, IMO. Marskell 16:38, 13 March 2007 (UTC)

Youtube

The discussion was previously at Reliable sources, now putting this here for lack of a better place: US$1 billion Youtube lawsuit. SandyGeorgia (Talk) 16:28, 13 March 2007 (UTC)

We're you going to add this to the article, or what? Alex43223 ^{T | C | E} 01:39, 14 March 2007 (UTC)

I don't actually have any idea what to do with it, or what to make of it. SandyGeorgia (Talk) 01:41, 14 March 2007 (UTC)

Contradiction

Without reading to much into both policies... correct me if I am wrong but does this policy's first sentence contradict Wikipedia:Article inclusion. Specifically: (to enter examples here)

• WT:A states: "The threshold for inclusion in Wikipedia is whether material is attributable to a reliable published source, not whether it is true."
• WP:AI states: "...just because something is true and verifiable doesn't automatically mean it should be included in Wikipedia."

Any comments? --FR Soliloquy 16:46, 13 March 2007 (UTC)

Here's a comment: Article inclusion is a personal essay and should not become a policy or guideline. Marskell 16:55, 13 March 2007 (UTC)

No conflict... AI talks about whether we should have an article on a topic or not ... this policy comes into play after that decision has been made, and talks about what can and can not be included in that article. For example... AI says that we might choose to not have an article on "Hairstyles of US Presidents", even if there are several books on the subject of Presidential hairstyles. The topic may not be considered notable enough for an encyclopedia. However, if the decision is that we should have an article on Presidential hairstyles, then ATT states that we must avoid original research and back statements up with reliable sources. In other words, AI talks about whether to have an article, while ATT talks about what goes into the article if it is written. Blueboar 17:05, 13 March 2007 (UTC)

Actually, if there are enough reliable sources on Hairstyles of US Presidents, then we should have an article on it. Wikipedia does not determine what is important or worthy of being included in an encyclopedia -- that job is left to reliable sources independent of Wikipedia and the subject itself.

I think AI needs to be reworded here. Perhaps what they're driving at are trivial details that may be verifiable, yet have not been noticed by reliable sources. For example, "Peter Griffin farted 15 minutes, 23 seconds into Episode 6 of Season 3 of Family Guy" may be verifiable (by watching said episode), but no reliable source is going document that. – Þ  01:37, 14 March 2007 (UTC)

I agree with Blueboar that there's no contradiction. Nothing in WP:ATT says that every single verifiable piece of information must be included in Wikipedia. It simply says information that is not verifiable should not be included. —Celithemis 01:50, 14 March 2007 (UTC)

Thank you for your comments. In short, there is no real contradiction and these two rules compliment each other! Though in the long run it may be nice to work WP:AI into WT:A, one is a only guideline while this WT:A is a policy. So I would probably be correct to say that the probability of doing such a thing is slim, right? --FR Soliloquy 03:58, 14 March 2007 (UTC)

WP:AI isn't a guideline, just a proposal that may or may not ever achieve consensus. —Celithemis 06:26, 14 March 2007 (UTC)

Just to chime in, WP:AI is a recent proposal that's still very much in the planning and editing stages right now. Any apparent contradictions will be ironed out, the intent is to make it complementary to this as a viable replacement to WP:N. Input, as always, is very welcome. --badlydrawnjeff talk 22:03, 14 March 2007 (UTC)

WP:NOR

Can you give your opinions about application of NOR in Talk:White_people#Behnam.27s_view? Basically he says that the sources must explicitly mention "white people" to be included in the article. And I say info about European people may also be given because that info is relevant to the topic as long as we do not make a synthesis (ex: because X is true for Europeans, X is also true for whites). The relevancy of Europeans to the topic has also been sourced by many sources throught the article. Lukas19 17:58, 13 March 2007 (UTC)

From WP:NOR ... "The only way to demonstrate that material is not original research is to cite reliable sources that provide information directly related to the topic of the article, and to adhere to what those sources say." The genetic studies Lukas19 is trying to add weren't about white people, or even used to bolster the 'genetic' view of white people. But Lukas19 is using them to add to the 'genetics' view. Hence, OR. Also, if you do investigate, try taking a look at the whole page rather than just one section. We could use some help. Thanks. The Behnam 02:14, 14 March 2007 (UTC)

As I told you many times before, "directly related to" does NOT equal to "having the same topic". It exactly means that, directly related. European people and white people are related topics. This has been sourced.Lukas19 05:41, 14 March 2007 (UTC)

Is this Original Research?

Could someone take a look at the article Unholy Alliance. My take is that there is a lot of original research and synthisis involved, but I need a second oppinion. Also, the article uses a google search result as a source... I don't think that qualifies as a reliable source. Perhaps one of the pages that came up IN the search might be a good source, but not the search itself. Blueboar 18:20, 13 March 2007 (UTC)

never mind... it's being nominated for deletion on exactly those grounds. I consider my querry answered.Blueboar 19:14, 13 March 2007 (UTC)

I'm wondering if we need to specifically point somewhere else for queries of this sort. (Did I just spell "queries" right?—I hate that word). Wikipedia talk:Attribution and Wikipedia talk:Neutral point of view etc. are meant to discuss the policies themselves, not specific cases. Not to discourage you Blueboar, but on a talk page that is rapidly becoming unreadable due to it's length, we should point things like this elsewhere. Marskell 21:57, 13 March 2007 (UTC)

Attribution Noticeboard? WAS 4.250 23:18, 13 March 2007 (UTC)

Attribution Talk (Policy) and Attribution Talk (Application)? I'd also really really appreciate if any of you can comment about the above section. Lukas19 05:43, 14 March 2007 (UTC)

We used to deal with things like this on Wikipedia talk:Reliable sources. Also, examples are a good way of considering how policies/guidelines work practically, furthering our understanding of the spirit of the policy/guideline, and helping us reconsider if it is properly worded to handle as many different circumstances as possible so that minimal use of WP:IAR will be needed. — Armed Blowfish (talk|mail) 14:52, 14 March 2007 (UTC)

I definitely think it would be useful to have a talk page that is dedicated to answering querries. At WP:RS we used to get at least two or three "is this reliable?" querries a week. I expect that before too long similar querries will be directed here. (Note: often the answer to these querries was, "Yes, but..." The majority of these types of questions were about reliably sourced statements of opinion, as opposed to statements of fact... and could be answered by pointing out the distinction between these two conscepts. We might save ourselves some hastle by including a statement about that distinction on our FAQ page). Blueboar 14:21, 15 March 2007 (UTC)

Congratulations!

I never thought I'd return here to see that Wikipedia:Attribution had finally made it. I thought the arguments from the past year or so would still be continuing in a never-ending cycle. Which archive should I look at so see how it finally came to fruition? Carcharoth 12:36, 15 March 2007 (UTC)

Stealing attributions from somewhere else.

A serious but common sin is to copy an attribution from a third party without look at it. For example, to copy a newspaper reference off a random web page without looking at the newspaper or mentioning the web page in the edit. This has long been discouraged as a guideline on WP:CITE, but I think it is very relevant to this page and ought to appear here as a policy. For convenience, here is the present text from WP:CITE:

It is improper to copy a citation from an intermediate source without making clear that you saw only that intermediate source. For example, you might find some information on a web page which says it comes from a certain book. Unless you look at the book yourself to check that the information is there, your reference is really the web page, which is what you must cite. The credibility of your article rests on the credibility of the web page, as well as the book, and your article must make that clear.

I propose to add the same or similar to this page. (I'd change "your article" at the end to "your citation".) --Zero^talk 13:09, 15 March 2007 (UTC)

This sounds like something that should be mentioned in the FAQ/advice guideline. Blueboar 13:23, 15 March 2007 (UTC)

My point is that it should not be just a guideline; it should be an enforcable policy. Violations of this rule are a serious problem which is hard enough to deal with without people claiming the rule is optional. --Zero^talk 14:44, 15 March 2007 (UTC)

It already is "enforcable policy". The FAQ explains enforcable policy with examples and details. This information explains what we mean by "source". If you did not look at it, then it is not your source and must not be listed here as the source used for the claim being sourced. WAS 4.250 15:08, 15 March 2007 (UTC)

Protect from moves

It seems sensible to protect this page from pages move vandalism [7], given that's an official policy page. ChazBeckett 13:27, 15 March 2007 (UTC)

Thank you... please do. Blueboar 13:29, 15 March 2007 (UTC)

Oh, I'm not an admin, I was just making a suggestion. ChazBeckett 13:30, 15 March 2007 (UTC)

references to subscription-only websites

I understood that references to Websites that are subscription only are not acceptable; it's been suggested at Talk:John T. Reed (regarding the persistent addition of unsubstantiated claims) that this is no longer the case, and this page was cited. I can't see that the issue is adddressed here; are there any opinions? Perhaps interested parties could also have a look at John T. Reed. --Mel Etitis (Talk) 17:12, 15 March 2007 (UTC)

It appears that what's being cited is the court case; the link is a "convenience link". Is this correct? --NE2 17:45, 15 March 2007 (UTC)

Subscription web sites are allowed; if we were going to ban information sources that have to be paid for, we would have to ban books and magazines too. However, information sources must be cited in the article. It isn't good enough to wikilink to another article, which in turn cites references, nor is it acceptable to depend on citations contained in external web cites. Information sources mentioned on the article's talk page don't count either. In summary, we don't need to discuss subscription web sites because the article does not cite any subscription web sites. --Gerry Ashton 18:06, 15 March 2007 (UTC)

I disagree with Gerry's argument that subscription sites are equivalent to having to purchase a book or magazine ... for almost all books and magazines you have the choice to obtain them at a library for free, something you can not do for the information on subscription sites. The rest of what Gerry says I do agree with. Blueboar 18:53, 15 March 2007 (UTC)

The citation isn't being given at all; it was mentioned in a Talk-page comment. The editor concerned steadfastly refuses to give a citation.

Books and magazines are available in libraries, so it's not clear that the analogy between them and subscription-only sites is justified. Certainly, Wikipedia:External links#Links normally to be avoided includes: "Links to sites that require payment or registration to view the relevant content." --Mel Etitis (Talk) 18:38, 15 March 2007 (UTC)

I think that to use an exclusive source which requires subscription would be problematic, as the verification cost for a reader could in principle be prohibitive. I would only use it if the cost is nominal and/or the information is supplementary and/or non-controversial. I agree that most books and magazines are in principle available for free in libraries. Crum375 18:45, 15 March 2007 (UTC)

Some readers will not be able to access to some sources no matter what we do. For example, young readers may not be allowed to go to the library, except on rare occasions, for safety reasons. Sources should not be prohibited just because they are expensive (unless the cost is kept artificially high to intentionally limit access to the source). --Gerry Ashton 19:00, 15 March 2007 (UTC)

Certainly the External links guideline indicates that subscription-only sites should normally be avoided, but the References and citation section of that guideline make it clear that it does not apply to references. As for books and magazines being available at libraries, some web subscriptions are available at libraries too, although I would guess the chances of getting free access to a magazine or book is better than the chances of getting free access to a subscription web site. --Gerry Ashton 18:53, 15 March 2007 (UTC)

We should also note that there is a difference between sites where you have to register, and sites where you have to PAY. I would be inclined to accept the first, but not the second. Blueboar 18:56, 15 March 2007 (UTC)

Just another opinion, seeing as this one seems to be debated: read the policy. Nowhere does it say that the source you are citing must be available without paying. It says it must be published, and must be reliable. These are the only criteria. JulesH 19:19, 15 March 2007 (UTC)

I second that opinion. The sources in the article in question are all freely publicly available from appropriate libraries: two newspaper articles and a court case. It is only online access that (might) require payment, and there is no requirement that sources be freely verifiable online. I have recently seen another article where someone claimed that only free online sources are acceptable... no way. The New York Times is a reliable published source even though its online archive is not free.

This is extremely relevant to science articles. We cite published journal articles that are freely available to the public at any research library (just pick up the bound journal off the shelf). Online access to these papers would require payment or subscription, however. CMummert · talk 19:26, 15 March 2007 (UTC)

Agree, thats what was said a few months ago last time this came up (anyone want to archive search..--BozMo talk 20:14, 15 March 2007 (UTC).)

It's difficult to follow this, as people keep conflating talk about the principle with talk about the particular case. If at John T. Reed a reference could be given to a source freely available in libraries, such as newspapers or magazines, then fine — my current question here would be irrelevant to that. (No reference of any sort has in fact been given, so the question is moot.)

If the only source given is one that requires a paid subscription, then (unless Wikipedia is willing to pay for such subscriptions) this essentially means that we have to take such citations on faith — and I don't see the difference in practice between that and saying that we take the claim itself on faith.

I should add that appearing on a private Website, available only to paying subscribers, is not the same as being published (i.e., it isn't being made public). A book might have to be paid for, but it is then open to anyone (not only libraries) to lend it to friends, etc. Moreover, it can usually be looked at in shops before or without purchase. None of these things is true of a paid-subscription-only Website, and I think that it shouldn't count as a verifiable, published source. --Mel Etitis (Talk) 20:47, 15 March 2007 (UTC)

Hmm. Personally I think the payment part is a red-herring. I am old and losing my memory but last time it turned on verifiability, good faith and subjective nature of what was listed. For example there are plenty of out of print but not public domain books which are only available as scans on pay per view sites, or heavily advertised sites. We kind of have to accept them on quite a few bits of history. Equally a lot of newspaper archives become pay only once the article reaches a certain age. The claim about Reed's court case is in principle verifiable from newspaper etc. and there is no particular reason to doubt it. It is a purely factual description without any silly language. So it should probably go in with a fact tag asking for a proper reference. There are lots of local newspaper reports etc like [8] which make it look to me like this should be easy to substantiate--BozMo talk 21:07, 15 March 2007 (UTC)

Not to cut in front but that seems like a valid issue on a community level -- i.e. access to the subscription sources through your local library, regional library, university research library, etc. are huge issues both in support of the services (usually protecting them from home-access), and to provide just such services to "researchers." Also remember that specialized and even private libraries do exist. Hell -- get the subscriptions and give it a go. In my little town that shall remain unnamed there are perfectly good Lexis-Nexus terminals sitting vacant 8 hours a day for months in the public Law Library near the courthouse, while a block away people queue up six and ten deep for a half hour slot on MySpace in the regular library. Xgenei 23:39, 17 March 2007 (UTC)

Tags

I added the "tags" subsection, copying its text from {{Citequote}}. I did this without previous discussion, because I thought this change is hardly controversial. But I was mistaken.

SlimVirgin reverted it saying, "keep it simple". I have to respectfully disagree. Simplicity cannot have priority before the completeness of instructions. I was not aware of this diversity of the tags, and when seing them, I immediately appreciated their utility. I have no doubts that I am not alone in this, both ignorance and appreciation. IMFO such things must be kept in one place, and the policy is the natural place for it. I understand the concern of SlimVirgin (in fact, I am thinking along the same lines, see my another section below). But "simplicity" for complex things is achieved by the inverted pyramid principle, rather than by merciless trimming.

Below is my suggestion for the "Tags" subsection:

• If you encounter a harmless statement that lacks attribution, you can tag it with the {{fact}} template (shorter synonym: {{cn}}), or move it to the article's talk page with a comment requesting attribution.

• {{Request quote}}, tagging a request for quoting inaccessible source, used for requesting a direct quote from the cited source for verification
• {{Citequote}}, for tagging quotations that are used without a citation
• {{Verify source}}, tagging a request for source verification, used for information that is doubtful or appears false.
• {{Verify credibility}}, tagging a request for source verification, used for information that is doubtful or appears false.
• {{Citecheck}}, popping up a box saying an article or section may have inappropriate or misinterpreted citations
• {{Not verified}}, popping up a box saying an article or section has not been verified and may not be reliable
• {{Unreferenced}}, popping up a box saying an article or section has no citation or reference for its information

Please comment. `'mikka 19:39, 15 March 2007 (UTC)

Most of them seem unnecessary; hard for editors to remember; some are saying the same thing. SlimVirgin ^(talk) 19:46, 15 March 2007 (UTC)

Also some of them have no method in place to resolve them. There is no "verification" process, so presumably every article qualifies for {{not verified}}; this would be true except that {{not verified}} is really a synonym for {{unreferenced}}.

The strangest one is {{citecheck}} - the description says that if you find the sources are not correctly cited, instead of moving the text to a talk page or deleting it you can use this tag - huh? The {{Failed verification}} tag is similary self contradictory - why tag a bad citation instead of just replacing it with {{fact}} or removing the unsourced claim? CMummert · talk 20:10, 15 March 2007 (UTC)

Y'all may be right. In this case, please arrange the cleanup of these tags, merge/delete/edit, whatever. But whatever is left, must be described in this policy, can we agree at least with this? `'mikka 21:32, 15 March 2007 (UTC)

We can't include in this policy a reference to every template any Wikipedian might create at any point. SlimVirgin ^(talk) 22:06, 15 March 2007 (UTC)

Why do you think the tags must be described in the policy? The policy is here to specify rules about article construction, not give advice to editors about how to follow the rules. I think there's an argument that they should be described in the FAQ, but not the policy itself. JulesH 23:35, 15 March 2007 (UTC)

I think that there is a difference between content-based tags and policy-based tags. Just as creation and editing of policies follow much stringent tules, the same goes with tags. There should no arbitrary proliferation of policy tags (see some criticism in this section above) and they must be described in one place, to ensure proper usage and to avoid confusion. `'mikka 15:19, 16 March 2007 (UTC)

Jules and Slim made good points. There is no need to list these tl's. ≈ jossi ≈ (talk) 01:37, 16 March 2007 (UTC)

Disagreed. `'mikka 15:19, 16 March 2007 (UTC)

Excess verbosity

Having pairs of subsections with synonymout sitles looks confusing to me (like, "No Original Research" and "Wikipedia does not publish original research or original thought")

IMO the section "Key principles" must be trimmed to the very nutshell and refer to detailed explanations in the subsequent sections with synonymous titles.

Alternatively, I'd suggest a more logical subsectioning:

• Key principles
  • "No Original Research": Wikipedia does not publish original research or original thought
    • Details
  • "Reliable Sources": Wikipedia articles must be based on reliable sources
    • Details

Any thoughts? `'mikka 19:29, 15 March 2007 (UTC)

This version went through about five months of discussion by hundreds of editors before being accepted, Mikkalai. I don't see the advantage of opening that up again, to be honest. SlimVirgin ^(talk) 19:31, 15 March 2007 (UTC)

I agree with Slim; the current version is comprehensive, yet succinct. — Deckiller 19:32, 15 March 2007 (UTC)

Respectfuly disdagree. My very point that it is not succinct: the structure is confusing and repetitive. "Hundreds of editors" is not a valid argument. Please address the merits of the proposal, not intimidate me with the head count. `'mikka 19:42, 15 March 2007 (UTC)

The merits are that it follows inverted pyramid structure, which is useful because it aids understanding by putting important points first then elaborating on the details of them. Trimming the key principles section any further is likely to make it incomprehensible or misleading. Restructuring as you suggest means we lose the advantage of having a key principles section. JulesH 23:31, 15 March 2007 (UTC)

Agreed. `'mikka 15:20, 16 March 2007 (UTC)

Question on an RS

If a source that is extensively used in an article, was deemed to be non-notable enough to be deleted from WP, would it still be acceptable as an RS? Torturous Devastating Cudgel 20:04, 15 March 2007 (UTC)

That's a leading question. There are enough unknowns in the circumstances you describe to mean yes or no are both possible. --BozMo talk 20:12, 15 March 2007 (UTC)

Well, I would rather not get into specifics, as it could provoke a firestorm, but the individual wrote two books, one is in 84 libraries, the other in 35. Google scholar lists one publication citing the author exactly once and the AfD was nearly unanimous. Oh, and I almost forgot to mention that the individual is a member of Scholars for 9/11 Truth. Torturous Devastating Cudgel 20:28, 15 March 2007 (UTC)

I think we'd need the specifics. Whether he's a reliable source depends on who he is, what he's writing about, and who has published him. It's unconnected to whether he's notable enough for a Wikipedia article. SlimVirgin ^(talk) 21:07, 15 March 2007 (UTC)

The specifics: Daniele Ganser and he is used in dozens of articles as an authority. Torturous Devastating Cudgel 23:35, 15 March 2007 (UTC)

Ah yes... 9/11 theory ... very tricky and highly controvercial. The quick answer is that it might be reliable and it might not, depending on the topic of the article and how it is being used. How to deal with various theories, books and websites centered on 9/11 is an issue that has plagued the various 9/11 related pages since at least 9/12. The relevant project has developed a host of consensus internal guidelines as to what is acceptable and under what circumstances... I would suggest you ask on the those project pages. Blueboar 12:28, 16 March 2007 (UTC)

Well, thats another thing, I only mention that as an adjunct although it is very loosely related to the material in question. Torturous Devastating Cudgel 13:02, 16 March 2007 (UTC)

Looking at the list of books cited during that AfD debate, and considering only those that are in English, I'd say:

• Nato's Secret Armies was published by a well-known academic publisher, and can therefore be used as a reliable source in many situations
• Reckless Gamble was published by a little-known small press, but nonetheless an academic one that at least claims to perform reviews for accuracy; it could be used, cautiously, in some situations

Looking over his newspaper articles, it seems most (if not all) of them are in small local-distribution papers, which suggests to me that including them in nationally important topics would probably be inappropriate. That's not a hard-and-fast rule, but it seems a sensible one. JulesH 13:44, 16 March 2007 (UTC)

Student newspapers

Student newspapers are not reliable; correct?--Sefringle 04:03, 15 March 2007 (UTC)

It would depend what they were being used to support. Generally, they're best avoided for anything contentious or for BLPs. It would also depend which newspaper it was. Some are very good, some not. SlimVirgin ^(talk) 22:15, 15 March 2007 (UTC)

Definitely a grey area. They generally have useful editorial control, with at least some degree of fact checking going on. But often it isn't good fact checking. They meet the minimum requirements, but obviously that only means they're valid for the least contentious of uses (which you're likely to be able to find other sources for anyway). JulesH 23:33, 15 March 2007 (UTC)

Is the The National Inquirer good? It all depends on the context! Is global warming actually happening... well according to All Gore all the scientist's reports say YES! But ask some news papers and many are skeptical and some deny it. Is it mean it's no good. I think you just have to maintain a good balance. --FR Soliloquy 04:30, 16 March 2007 (UTC)

As the policy says, sources may be reliable in one contexts and not in others. IMO a newspaper is good as a source of news firsthand reported by their reporters, or opinions of reputable persons interviewed, or articles signed by experts in the domain. Any other information coming from a newspaper, if it is not attributed in it, is not reliable. Clearly, newspaper editors and reporters are not experts in math or physics, etc. The same applies to student newspapers: you must further look at the possible domain of its expertise. Clearly they are supposed to be experts in local school news, but any other info coming from them unattributed or not written by an expert in the relevant domain is dubious IMO. `'mikka 15:28, 16 March 2007 (UTC)

Concur with the gist of all of this. A source is not necessarily unreliably just because it is a student newspaper. — SMcCandlish [talk] [contrib] ツ 16:30, 17 March 2007 (UTC)

OR wording seems to go too far and to be self-contradictory

The WP:ATT version of OR is pretty unclear and seems a bit excessive, at least in its more stringent interpretations. By my reading it appears to effectively ban the giving of examples that someone else notable didn't write, unless they are mathematical examples (meanwhile using other people's previously-published examples more than very minimally and with direct citation right at the point of usage would be a copyright violation under some jurisdictions' intellectual properly schemes).

For instance, '...an example of this type of redundancy is found in the phrase "get the cat off of the table", since the word "of" can be omitted and the resulting phrase, "get the cat off the table" has the same meaning.' Such simple examples are crucial for many articles, especially those on language concepts, yet the new WP:OR appears to label them a novel synthesis or new theory, when really they are utterly straightforward logical deductions from incontrovertible facts (which can be sourced, of course). Basically a strong argument can be made either way that this is basic logic and simple writing, quite appropriate for an article here, or that it is forbidden, uncyclopedic OR because the example itself didn't come from a journal or book, however closely it may follow similar examples in such sources. Given the "deletion fever" running high on AfD, I think OR's vagueness and seeming self-contradition - "no you can't come to any conclusions, oh, wait, actually yes you can, but only if they're basic logic, which we don't define here, or math" - is pretty likely to lead to inappropriate deletions of articles.

This section needs clarification.

— SMcCandlish [talk] [contrib] ツ 21:32, 16 March 2007 (UTC)

PS: The fact that the FAQ only has one annotation with regard to OR's meaning and applicability is a bad sign, and suggests that the entire concept needs some work before it will be of true value as a guideline. I imagine the talk page of the original WP:OR can provide a lot of FAQ fodder. — SMcCandlish [talk] [contrib] ツ 21:47, 16 March 2007 (UTC)

I see what you mean. The problematic phrasing, as I see the issue, is the inclusion of previously unpublished "explanations", which would probably encompass examples. I think this can be removed without much issue. The word was added only a couple of weeks ago (here), after the policy had consensus. I don't see any discussion about the change, and don't see why it was needed, so I'm going to go ahead and revert it. JulesH 22:46, 16 March 2007 (UTC)

Actually, I think this change was particularly wrong, and should have been caught sooner. The entire purpose of wikipedia, what we strive to achieve here is to write explanations in our own words of information other people have published. If we're forbidden from introducing new explanations, this goal is clearly unachievable. JulesH 22:55, 16 March 2007 (UTC)

As I noted in my edit summary, I can see how an unsourced 'explanation' would be used by some editor to advance a position, clearly violating OR. I would keep the current language, unless a consensus appears otherwise. Crum375 22:56, 16 March 2007 (UTC)

I don't see a problem in adding simple explanations in clearly non-contentious issues, when not 'advancing a position'. Crum375 22:59, 16 March 2007 (UTC)

Policies must not forbid writing that is sometimes inappropriate, only writing that is always inappropriate. Furthermore, the part of the policy that has just been edited does not say anyting about advancing a position. --Gerry Ashton 23:02, 16 March 2007 (UTC)

I agree about the non-contentious cases. I guess if there is a contentious case, where an example is used to advance a position, it could be shot down using NPOV and/or viewed as an 'interpretation', so it would be covered. BTW, 'synthesis' does use the words 'advancing a position'; I would think that 'creating an example to advance a position' is no less OR-ish. Crum375 23:18, 16 March 2007 (UTC)

Just to get my !vote in: I believe we all understand the point that "explanation" can be used to mean theory (or interpretation, etc.), but this does not mean that the word "explanation" needs to be placed next to "theory". It's just intuitively obvious. If some dork-o tries to defend a nonsense POV article in AfD with "but, I'm just explaining...", no one is going to buy that. Insisting on treating the word "explanation" like "theory" here has severe negative consequences because of the deep abiguity it introduces. And I have to say Gerry Ashton's point just above hits the nail right on the nose. One of the smartest things I've seen said on WP in a week or so. His second point is also right on the money - this is WP:ATT not WP:NPOV. I know some people here feel a deep need that I do not understand to collapse separate policies and guidelines into longer, more (supposedly) melded ones, but enough is enough. Anyway, as I said in edit summary, JulesH is utterly right on this. Every article on Wikipedia is a "new explanation"; if it wasn't it would be a blatant copyvio. — SMcCandlish [talk] [contrib] ツ 01:31, 17 March 2007 (UTC)

I don't buy JulesH's reasoning, though I respect the intention. "Theory" has different - quite distinct - uses and is not a synonym for explanation. Explanation should be included if we are talking about no original explanations being permitted (and it was in the old NOR policy). JulesH's reasoning can apply to synthesis or interpretation - of course Wikipedia, like other encylopedias, provides syntheses, and interpretations, and explanations - but our policy is that these must come from attributable sources and nto be original to the editor. I don't see any controversy about keeping "explanations" when we are clear the point is whether or not they are original or attributable. Slrubenstein | Talk 16:49, 19 March 2007 (UTC)

What Slrubenstein says makes sense to me. ^High_InBC^{(Need help? Ask me)} 16:53, 19 March 2007 (UTC)

I can't concur on this at all, but I'm not going to edit war over it. — SMcCandlish [talk] [contrib] ツ 17:33, 19 March 2007 (UTC)

The key difference is that an explanation is a much lower level textual detail than a synthesis (which refers to how different pieces of information are combined) or an interpretation (which refers to the meaning derived from the information). Explanations would tend to be on a much lower level: a set of points that together can be used to understand the meaning of something. It's not quite on the level of words, but not far off. JulesH 19:37, 19 March 2007 (UTC)

The problem reported in this topic remains unresolved since the fix for it was reverted on the basis that policy "used to" say "explanation", so it still should. I don't find that reasoning compelling in any way. — SMcCandlish [talk] [contrib] ツ 17:33, 19 March 2007 (UTC)

Are examples explainations?

One of the arguments about whether original explainations are allowed is whether examples constitute original explainations. For example, if I take material from two different sources, and both sources give examples, for the sake of conciseness and smooth flow, I might want to modifiy one of the examples to use one example throughout the article. If I'm writing an article about electricity, and one source uses a 6 V battery example while discussing resistance, while a second uses a 12 V battery while discussing current, I might want to modify the first example to use 12 V in order to make the article flow more smoothly. In my view, the revision of the values would be an elementary calculation. So I would like to know if Slrubenstein considers original or modified examples to be original explainations. --Gerry Ashton 18:54, 19 March 2007 (UTC)

I think that examples can be a form of explanation, and in my opinion any example that is not based on reliably published sources could be viewed as OR. I would probably allow simple examples for non-contentious topics, like the one you mention, assuming they otherwise conform to published knowledge, but I would shy away from them, like any original interpretation or analysis, in contentious cases where used to advance a position. Crum375 19:40, 19 March 2007 (UTC)

Would you, therefore, support a change that restricts original explanations to those that do not advance a position that cannot be attributed to a reliable source? This would be different from the current text in that it doesn't require the explanation itself to have been published by a reliable source, only the points it is trying to make. JulesH 19:44, 19 March 2007 (UTC)

This is the current text: "Material counts as original research if it: ... introduces an analysis, synthesis, explanation or interpretation of published facts, opinions, or arguments without attributing that analysis, synthesis, explanation or interpretation to a reliable source who has published the material in relation to the topic of the article". That sounds fine to me. Crum375 19:50, 19 March 2007 (UTC)

What if a series of examples is added to an article, to advance the position that 'one side' in a public controverly uses illegitimate tactics? For example, the Astroturfing article contains predominantly examples of accusations that various "pro-industry" PR campaigns are masquerading as grassroots initiatives. Is this being used to advance the POV that "industry lies and cheats" while "Environmentalists are good and honest"?

How many 'examples' are needed to explain a concept? And when does this cross over into being a list of offenses (used to demonize a side) rather than a set of "examples used to illustrate the idea"? --Uncle Ed 19:56, 19 March 2007 (UTC)

If your 'examples' are reliably published, they would be acceptable, if pertinent. Crum375 20:00, 19 March 2007 (UTC)

Consensus

It is important that anyone who disagrees with the proposed change be willing to discuss their objections here. SlimVirgin and Crum375 have reverted the proposed sentence, against the consensus on this talk page, but have not articulated here any reason for preferring the version without that sentence. We still haven't heard here, if I remember right, from anyone who supports the new, unqualified "not whether it is true" wording an example of a situation in which they think that wording would have a beneficial effect in comparison to not having those words.

Please do not revert the new sentence without expressing clearly here the reasons for opposing it. We're still in the process of trying to find a wording that consensus can be built around after the new, unqualified "not whether it is true" wording was proposed. We need everyone involved to state their opinion clearly so that we can search for a wording that everyone is reasonably satisfied with. Reasons have been stated for including the sentence "Not everything which is attributable is worhty of inclusion." No reasons have been stated here for not including this sentence.

I'm sorry, but I'm not willing to interpret silence as objection. If people have objections, they need to communicate not only the fact that they have an objection, but also the reasoning behind the objection, so that a discussion can continue from their in search of a wording that everyone can live with. I've noticed that some changes have been made recently to the policy without my having commented in advance, so I don't think taking silence as objection is the norm. We also need to consider the silence of those who used WP:Verifiability for many months, without seeing any merge tag on it, only a notice that it was policy. If such people are now reverting the new wording in various ways it should not be surprising. --Coppertwig 23:34, 16 March 2007 (UTC)

Indeed. I don't even want to get started on the hijacking aspects; just not going to go there right now. To be honest, I'm smelling a whiff of WP:OWN around here. The revert rate is high and rather testy (or completely silent, which is even worse), even for a guideline/policy page, and often seems to be based on "just because I said so" rather than adequately explained reasons; I'm not just talking about today in particular and not singling anyone out; it's just something I've observed as a general, multi-party pattern ever since the merge of everything under the sun into WP:ATT (which I frankly think was an enormous mistake, but I and many, many others weren't even aware of this consolidation move. D'oh. I said I wasn't going to go there...) Anyway, as to editwarring, WP:N was like this in Nov. - Dec. 2006 and part of Jan. 2007, and the fights this over-control caused were really something (and not something good). That didn't get to be a rational guideline that had actual, cognizable consensus until around early Feb. 2007. Anyway, I agree vehemently that silence is not objection, with the caveat that silence for 5 minutes or even 5 hours doesn't mean anything at all. Silence for several days, however, can be taken as assent. If it turns out to not be assent, but just lack of attention, failure to understand the proposal, or considering it so silly as to not have been worth commenting on, that should become pretty apparent, and frankly shouldn't be very common. If it is common, then this means that the editors doing the testy reverts need to remember how to use a talk page.

All that said, I don't agree with the inclusion of the "not everything which is attributable is worthy of inclusion" bit. This concept is already covered elsewhere, and this policy for whatever it is presently worth or will be worth at some point, is not about overinclusion of irrelevant but true facts, it is about the requirements for inclusion at all. Completely different topics. That said, I think the point is a good one, and if it is not being made clearly enough where it should be made, then that situation should be improved. But as for this particular policypage, "no reasons have been stated here for not including this sentence" is no longer true. Sorry; nothing personal.— SMcCandlish [talk] [contrib] ツ 01:31, 17 March 2007 (UTC)

Synthesis needs its own bulletpoint

Per the recent edits of "What is Original Research?" we now have the following bulletpoint:

"[Material counts as original research if it] introduces an analysis, synthesis, or interpretation of published facts, opinions, or arguments without attributing that analysis, synthesis, or interpretation to a reliable source who has published the material in relation to the topic of the article."

While this statement is perfectly in line with the policy regarding analysis and interpretation, it is not representative when it comes to synthesis. Take the dicdef of wikt:synthesis:

"the formation of something complex or coherent by combining simpler things"

Or to simplify, any collection of numerous things, in this case sources. When you consider this definition in regards to the policy, this means that the new bulletpoint effectively states that you cannot use any combination of sources in an article unless some other reliable source has used that exact same combination of sources previously. This is obviously wrong; Wikipedia is at its very nature a collection of information gathered from multiple sources. It is only when that combination of sources advances some position or induces the reader to draw a conclusion that it becomes OR.

To remedy this problem, I propose the above bulletpoint be expanded into the following two bulletpoints (or similar):

• introduces an analysis or interpretation of published facts, opinions, or arguments without attributing that analysis or interpretation to a reliable source who has published the material in relation to the topic of the article.
• introduces a synthesis of published facts, opinions, or arguments that serves to advance a position without attributing that synthesis to a reliable source who has published the material in relation to the topic of the article.

Thoughts? -- Y|yukichigai (_ramble argue ^check) 00:12, 17 March 2007 (UTC)

Concur. I think someone simply liked the word "synthesis" and wanted the text to include it, but didn't understand the ramifications of doing so. Also, the text badly needs a copy edit to get rid of the blah, blah or blah repetition. It's really tedious. — SMcCandlish [talk] [contrib] ツ 00:58, 17 March 2007 (UTC)

I also agree. To take this to an extreme, 1 + 1 = 2, and everyone agrees, but 1 + 1 + 1 = 3 and the current wording almost prohibits it. If the result is obvious and non-controverial, synthesis is OK. - Peregrine Fisher 07:21, 17 March 2007 (UTC)

Even 1 + 1 + 1 = 3 can be "original research", such as in a discussion of whether or not Christianity (father, son, holy ghost) is monotheistic or not. WAS 4.250 21:47, 18 March 2007 (UTC)

I see that at some point the section on synthesis has lost the text "in a way that advances a position favored by the editor" that was previously in it (e.g. in this version). Now, I can see why this text has been removed: it doesn't say what we want it to say. How do we know what position the editor favours?

How about replacing it with "that serves to advance a position that is not attributable to a published reliable source"? Alternatively, we could add something like "Any material that argues in favour of a position that can be attributed to a published reliable source is not original research" to the "What is not original research" section. Comments? JulesH 07:42, 17 March 2007 (UTC)

Just to clarify, the distinction between my first suggestion and yukichigai's is that it doesn't require attribution, but attributability. This is necessary in order to avoid contradicting the definition of something that is unattributable as OR, but not necessarily something that is unattributed. JulesH 07:45, 17 March 2007 (UTC)

Support. I think that works beautifully. — SMcCandlish [talk] [contrib] ツ 16:28, 17 March 2007 (UTC)

Aren't we splitting hairs here? ≈ jossi ≈ (talk) 00:04, 18 March 2007 (UTC)

Um, no. Been to AfD lately? Any way even vaguely conceivable to misinterpret or miscast policies and guidelines, pro or con deletion, will happen, again and again, and sometimes even build a weight of false precedent, until thwarted by this so-called "hair-splitting" (i.e. clarity). I'm not one to advocate WP:BEANS or WP:CREEP style true hairsplitting, of course, but the largest and deepest policy debates really are almost always over wording that seems very minor or barely any different to people not involved in the debate over that wording. This is hardly a Wiki thing either; you should try watching US legislation, for example, evolve day-to-day some time (and you can, too, at the THOMAS server). — SMcCandlish [talk] [contrib] ツ 00:15, 18 March 2007 (UTC)

Interpretation and application of policy is still within the domain of editor's responsibilities. We cannot craft policies to address each single instance of its application. "Hecha la ley, hecha la trampa", (Spanish: make a rule and you would have also have made a way to break it). The current formulation of "synthesis" as it applies to NOR is more than sufficient for this policy. If you want to expand on details and give examples, a better place would be the WP:ATT/FAQ. ≈ jossi ≈ (talk) 21:12, 18 March 2007 (UTC)

Mmm... that's non-responsive, really. The proposed change ("attributability") is emphatically not about "expand[ing] on details or giv[ing] examples", it is about defining; you're making straw man argument.  :-) — SMcCandlish [talk] [contrib] ツ 23:44, 18 March 2007 (UTC)

This isn't a case of expanding on details or giving examples: this is clarifying that the definition does not apply in cases where it should not. Original Research, as we've always defined it in the past, is (among other things) synthesizing multiple sources in order to make a point that hasn't been published by a reliable source. The problem is that the current text doesn't include the restriction to unpublished points. The definition has been broadened, and I don't think anyone really intended to broaden it. JulesH 21:20, 18 March 2007 (UTC)

I am readig Wikipedia:Attribution#No_original_research, JulesH, and cannot see what has been broadened. Care to help me understand what in your view has been broadened? ≈ jossi ≈ (talk) 21:32, 18 March 2007 (UTC)

Note the removal of the language "that advances a position favored by the editor"; removing this language is the problem. Because the word "synthesis" has been left the new bulletpoint effectively states (as I stated above, don't know how I can be any more clear) that ANY synthesis, any "formation of something complex or coherent by combining simpler things", is Original Research. This does encompass the kind of synthesis we want to avoid -- the kind where there are unpublished conclusions drawn (or directed to be drawn) -- but it also prohibits anything that uses a combination of sources when that exact combination of sources hasn't been used by some other source previously. Effectively, removing the "advances a position" language and leaving the word "synthesis" in has made all of wikipedia Original Research, since every wikipedia article is "the formation of something complex or coherent by combining simpler things", i.e. making a coherant article by gathering information from multiple sources. -- Y|yukichigai (_ramble argue ^check) 22:01, 18 March 2007 (UTC)

Entirely agreed, though I think we are now having two different conversations that need to either be tied together more clearly, or forked into separate topics. — SMcCandlish [talk] [contrib] ツ 23:44, 18 March 2007 (UTC)

Yukichigai elsewhere (on the FAQ page) has proposed replacing "synthesis (of)" with "analysis or interpretation(s) (of)". I'm strongly inclined to support that change. "Synthesis" is being misused here, and is already the subject to two disputatious threads on this talk page. — SMcCandlish [talk] [contrib] ツ 01:32, 19 March 2007 (UTC)

Contentious topics

This sentence sparked a question in my mind:

• Wikipedia articles should rely on reliable, published secondary sources wherever possible. This means that we only publish the opinions of reliable authors, and not the opinions of Wikipedians who have read and interpreted primary source material for themselves.

What about the case where secondary sources disagree with each other, and one group of secondary sources is very popular in the public media (or even among Wikipedia contributors)? I refer to cases where one side (frequently US liberals) claims that there is a "consensus" on any of many scientific questions.

Should Wikipedia agree with the Liberal POV here, and say that "there is no controversy ... there is a consensus ... etc."? Or would this be giving undue weight to one side in a debate?

How much of a majority must there be on a disputed topic, for the article to adopt one position on it as the "consensus position"? Is 80% to 85% considered consensus on scientific topics, as it is for afd and rfa votes? Would this mean that if only 10% to 15% of scientists in a particular field disagree with the mainstream, there views should be excluded from the article in question? (Or just marked as "minority views"?)

And if a particular viewpoint is deemed "in the minority", are editors required to delete any citations which support it? Would it violate 'undue weight' to cite a peer-reviewed scientific journal article which goes against the mainstream? If so, what percent of support among scientists must the mainstream view have, before all contrary views must (or may) be excluded from an article?

I would propose that 99.8% is certainly enough for that (i.e, 499 out of 500). But what if it's only 95% (i.e., 19 out of 20)? Where is the cutoff?

Or what if Wikipedia editors are unable to determine the percentage of support? Suppose all we know on some obscure (or deeply technical) matter, is that political advocates outside of Wikipedia are citing duelling studies? Passive smoking causes cancer! No, it doesn't! Antarctica is warming and the glaciers are melting! No, it's getting cooler and the ice is thicker than ever!

Do we just guess that it's 50-50? Or avoid estimating the percentage ourselves? Do we quote one side, which says "everyone agrees with me"? If we don't know, are we as editors allowed to leave it indeterminate? --Uncle Ed 00:17, 19 March 2007 (UTC)

Yes, it would be undue weight, and yes, we should avoid making up percentage estimates (blatant OR). — SMcCandlish [talk] [contrib] ツ 01:37, 19 March 2007 (UTC)

I'm not sure I see the link between the sentence of policy you quote and the question you ask. In fact, the question you ask should really be at WP:NPOV, not here. JulesH 08:19, 19 March 2007 (UTC)

I agree with JulesH. The above is not a WP:NOR issue. It's a WP:NPOV issue. J Readings 09:22, 19 March 2007 (UTC)

Let's try and give an answer, anyways. We shouldn't be trying to judge the percent of agreement and disagreement, first off. If there is a "consensus agreement," that should only be stated if some secondary source states that some view has a "consensus." If the New York Times says "scientists have reached a consensus that global warming is real," then there is a statement about consensus that we can use. We basically have to trust secondary sources, else, who can we trust? Our own opinions have no bearing on this. Undue weight is not a problem if peer reviewed scientific journal are cited, it's only a problem when they aren't. A peer reviewed journal is a good source, so we can use it all we want. Now let's look at the possible downside of all this. Say global warming is mentioned in 1000 articles; 900 say it's real, 100 say it isn't. One could write an article with 10 refs from each, making it seem like it's more disputed than it is. There's nothing these policy pages can really say about this; this is where we have to trust our editors. It's an imperfect world, and we all just do the best we can. If an editor uses the fact that there's lots of reliable sources for each viewpoint to push their POV, it just has to be worked out in the talk pages. It may seem like using math to answer these questions is a good idea, but really it's best done through human discussion. - Peregrine Fisher 09:52, 19 March 2007 (UTC)

Thanks for your response, Peregrine. I have a follow-up, too.

Whether JulesH and J Readings see the connection or not, the policy page says:

• Wikipedia:Attribution is one of Wikipedia's two core content policies. The other is Wikipedia:Neutral point of view. Jointly, these policies determine the type and quality of material that is acceptable in articles; that is, content on Wikipedia must be attributable and written from a neutral point of view. Because the policies are complementary, they should not be interpreted in isolation from one another.

So I hope you don't mind my asking a question that seems 'more' connected to NPOV. I'm trying to figure out how these 2 complementary policies relate.

Let me try to clarify my question, because it relates to a pattern of editing engaged in by many of contributors on dozens of topics (involving 100s of articles).

Often there is a dispute about a historical or scientific matter. Like, did the Holocaust occur? (Or, did events similar to Hitler's holocaust occur in other times and places?) There are many political controversies over the environment, like Alar, DDT, "greenhouse gases", etc.

Partisans in the 'real world' outside of Wikipedia cite sources to support their positions. We have pictures in Life magazine! No, those pictures were faked! China admitted killing 20 million! No, they never said this! DDT is dangerous to birds and people! No, it's safe enough to eat by the spoonful! (There are quotable secondary sources for each of these shouted examples.)

Enter Wikipedia. Since we have a policy on neutrality, readers would naturally expect to find verifiable sources which touch on these issues. All they have to do, to be really sure, is to "verify" the verifiable. E.g., follow the link and see if what the source really says - and the way the article summarized the source - are in agreement.

But what if an issue is so complex that it consists of multiple aspects, each of which is an issue in itself? It can be too much for average readers, and even too much for some highly educated ones. At some point, the temptation is to cite someone's evalution or assessment of the whole thing:

• According to (blank), scientists are unanimous about Alar (or DDT, etc.)

But what if sources which make assessments are in disagreement? Or what if they state that there is a consensus but don't show that there is one?

Or what if there is a question about credibility? Suppose a source does not only only make an assessment of the facts, but also recommends policy? Or what if a source is seen as (yes, that's a weasel word) beholden to financial or ideological interests? The group you cite accepts pro-industry funding! No, the group you cite accepts anti-industry funding! Your group is biased in favor of socialism! No, your group is biased against socialism!

It's not easy to determine whether a source is unbiased and objective. Is the EPA considered utterly impartial and completely independent of partisanship? How about the Atlanta CDC? Or agencies of the United Nations? What makes a source objective? The fact that it got broad international support from representatives appointed by governments?

One last question. Suppose prominent organizations run by governments or universities take a position on something; and they issue statements assessing a particular question; and these statements purport to present the "scientific consensus" on something. Suppose then that several individual scientists disagree (in public, or by publishing in peer-reviewed journals) on (1) the science itself and/or (2) whether there is a scientific consensus.

Do we:

1. accept the assessments of the prominent organizations as factually correct and reject all dissenting views?
2. list the prominent organizations which assert that there is a consensus, but refrain from endorsing these assertions; and also list the dissenting individual views?

Sorry this is long, but my question is about how NPOV and Attribution relate to each other on issues of historical or scientific controversy - particularly when sources assessing the matter disagree on "how much agreement or disagreement there is". --Uncle Ed 15:56, 19 March 2007 (UTC)

I think we try to "list the prominent organizations which assert that there is a consensus, but refrain from endorsing these assertions; and also list the dissenting individual views." All we really have is editor's judgement. It doesn't work perfectly, obviously, but it's the best we can do. - Peregrine Fisher 18:12, 19 March 2007 (UTC)

Remove silly instruction creep

"Be particularly careful when proponents of such claims say there is a conspiracy to silence them" is WP:CREEP to the point of WP:BEANS. The entire sentence should be removed. If the day comes when conspiracy theorist editors' paranoia becomes such a common issue, and such as serious one that other editors need to "be particularly careful" (or else what? Aliens will abduct them?), then maybe consider some language like this. — SMcCandlish [talk] [contrib] ツ 15:45, 19 March 2007 (UTC)

I agree, and anyway it's not clear what "be careful" means. Should contributors reject all 'claims of being silenced' out of hand? What if a scientific paper finally passes peer review after 3 years? Is this an indication of sloppy work on the part of the paper's authors? Extra scrutiny routinely applied to views which challenge the mainstream? Evidence that the paper is pseudoscience? Evidence that the journals are biased?

I don't think Wikipedia articles should try to answer those question, but instead refrain from trying to evaluate claims of 'truth vs. pseudoscience' and 'normal procedures vs. bias'.

On the other hand, proponents of novel views often spend more time complaining about suppression than they do explaining the merits of their views. So the caution you quoted above has some merit. Maybe the article should just (1) state what the proponents' views are and (2) report that they asserted there attempts made to silence them. --Uncle Ed 16:22, 19 March 2007 (UTC)

When I get some guy on the Earth article, claiming it is 6000 years old based on the bible, and he says I am trying to silence him, I just demand reliable sources that demonstrate this is believed by anything but a small minority. ^High_InBC^{(Need help? Ask me)} 16:11, 19 March 2007 (UTC)

That makes sense, but bear in mind that about 45% of Americans are Creationists - I'm not sure how many of those are Young Earthers, but there are anything but a tiny minority.

In the field of geology, however, I have YET to see anything published about the age of the earth other than the mainstream view, i.e., it's billions of years old. Am I missing something? --Uncle Ed 20:12, 19 March 2007 (UTC)

I'm not sure removing it is a good plan. The advice it gives is helpful: I've certainly seen editors claim a conspiracy to silence their views (the particular editor I'm thinking of was a well-published MD who claimed that his research against medical consensus -- suggesting that treating mental conditions with drugs was inappropriate -- was being suppressed by the pharmaceutical industry), and it has been a helpful tip-off that their contribution may not be entirely reliable. I think maybe it, and perhaps some of the other text from the same paragraph, would be better off in the FAQ, however. JulesH 19:51, 19 March 2007 (UTC)

Moving it to the FAQ would (at least for now) resolve my concern. — SMcCandlish [talk] [contrib] ツ 15:23, 20 March 2007 (UTC)

Original Research definition -- unified discussion (hopefully)

OK, this is the current definition of OR:

Material counts as original research if it:

• introduces a theory, method of solution, or any other original idea;
• defines or introduces new terms (neologisms), or provides new definitions of existing terms;
• introduces an argument without citing a reliable source who has made that argument in relation to the topic of the article; or
• introduces an analysis, synthesis, explanation or interpretation of published facts, opinions, or arguments that advances a point that cannot be attributed to a reliable source who has published the material in relation to the topic of the article.

(I've just made a change which I hope won't be reverted, which is to reintroduce the old language in the last bullet point that specifies that something only qualifies if it advances an unpublished point.)

This definition is, frankly, a mess. Taking it step by step, interpreting each bullet point as a single sentence:

Material counts as original research if it introduces a theory, method of solution, or any other original idea

This is just wrong. Huge numbers of pages on wikipedia introduce theories or methods of solution, and they aren't OR. The definition of "original" in the last clause is ambiguous; by at least one interpretation almost every page in wikipedia falls foul of this definition.

Material counts as original research if it defines or introduces new terms (neologisms), or provides new definitions of existing terms

Again, defining a term is not OR. Defining a term to mean something that it hasn't been defined to mean by a published reliable source is OR.

Material counts as original research if it introduces an argument without citing a reliable source who has made that argument in relation to the topic of the article

This isn't quite so bad, but note that it requires attribution, not attributability, which contradicts the text of the following paragraph.

My proposed fix to this is to substitute this text:

Material is original research if it introduces, in a way which promotes a point of view or idea that cannot be attributed to a published reliable source, any of the below:

• a theory, argument, idea, or definition, or
• a synthesis of information from multiple sources, or
• an analysis, explanation or interpretation of published facts, opinions or arguments.

Comments? JulesH 20:13, 19 March 2007 (UTC)

Strong support as far as it goes. This entire section has been messed up and confusing for a very long time. In particular, defining "explanation" as OR means every article is OR, since all Wikipedia does is explain things; that's the very function of an encyclopedia. I realize that sometimes one can use the term "explanation" when what one really means is "theory, argument, idea or definition", but no one is going to be fooled by this, so there is no reason to include "explanation" in the littany in the present version. I would go further and also remove "synthesis" for reasons already covered elsewhere; the word is simply being misused here. Every WP article that cites more than one source is a "synthesis" under the most common definition of that word. It might be possible to use the word in a different, more explained way, however, such as "* a synthesis of information from multiple sources that results in such a theory, argument idea or definition". The appearance of "example" in the reworded verison above isn't nearly as problematic as in the "live" version because of the new version's qualifying lead-in, so I'll hold off complaining about it. In short, I think this new version solves a great many problems, I can't think of any that it introduces, and to the extent that it might not solve every problem with the original wording, that can be addressed later as the need arises.

PS: I would change "below" to "following", and change commas to semicolons before the two final "or" instances.

— SMcCandlish [talk] [contrib] ツ 20:23, 19 March 2007 (UTC)

Actually, it occurs to me that a simpler definition is possible: "Material is original research if it promotes a point of view or idea that cannot be attributed to a published reliable source." The rest could then be relegated to explanations of things that are included, rather than being part of the definition. Comments on this would be appreciated, too. JulesH 20:14, 19 March 2007 (UTC)

Either way is fine by me. Both versions would be equally rational compared to the current self-contradictory mess. — SMcCandlish [talk] [contrib] ツ 20:24, 19 March 2007 (UTC)

You guys are on the right track. Thanks for taking the time to do this.

How about:

• introduces an a novel analysis, synthesis, explanation or interpretation of published facts, opinions, or arguments that advances a unattributed point

--Uncle Ed 20:49, 19 March 2007 (UTC)

Concur on the gist; I think keeping the term "novel" in there is actually important. But it would need to read:

• a novel analysis, explanation or interpretation of published facts, opinions or arguments

to go with the new version. — SMcCandlish [talk] [contrib] ツ 21:35, 19 March 2007 (UTC)

Okay, I think I've figured out this mental block over "synthesis"

I did a little checking, and I've figured out (I think) where this whole "we need to use the word 'synthesis'" thing is coming from. The original Wikipedia:No original research policy's wording was based heavily on several things written by your hero and mine, Jimmy Wales. In particular, let's look at this little snippet (emphasis mine) that was cited as the basis for the old "synthesis" section of WP:NOR:

Some who completely understand why Wikipedia ought not create novel theories of physics by citing the results of experiments and so on and synthesizing them into something new, may fail to see how the same thing applies to history.

— (Wales, Jimmy. "Original research", December 6, 2004)

Seems all well and good, but here's the important bit that everyone seems to have overlooked: the definition of synthesis is different than the definition of synthesize. One is not just the verb form of the other. The difference is subtle, so I'll explain it clearly: synthesis is any new collection or combination of "things" that may or may not change the nature of the "things", whereas the process of synthesizing something is making something new out of a number of other "things". To simplify, synthesize means to make something new, whereas synthesis is the mere form of putting things together.

As far as I can understand it, the problem is that numerous editors -- not to single folks out, but SlimVirgin and Jossi in particular -- are applying the definition of the verb to the definition of the noun. The two are not the same, hence my objections to the current way "synthesis" is being used. -- Y|yukichigai (_ramble argue ^check) 23:07, 19 March 2007 (UTC)

I think if you consult most dictionaries, as well as common usage, you'll discover that, quite logically, 'to synthesize' means 'to create a synthesis'. Your assertion that the verb 'synthesize' is somehow weaker than the noun 'synthesis' is quite novel, and in fact, unless attributed, WP:OR. Crum375 23:17, 19 March 2007 (UTC)

Do you not understand how to read? I provided not one but two links to dicdefs of synthesis and synthesize. Read them. -- Y|yukichigai (_ramble argue ^check) 23:22, 19 March 2007 (UTC)

Just to be sure, I looked up both words on dictionary.com. 11 entries for synthesis, 7 for synthesize. Only the intrasitive form of one definition is listed as meaning "to create a synthesis"; 5 of the 7 definitions specifically include the word "new". -- Y|yukichigai (_ramble argue ^check) 23:36, 19 March 2007 (UTC)

I can always improve my reading skills, but here is the primary definition from your two links:

• Synthesize: "to combine two or more things to produce a new, more complex product"
• Synthesis: "the formation of something complex or coherent by combining simpler things"

It sure seems to me that the verb creates the noun. Crum375 23:39, 19 March 2007 (UTC)

It sure does: Synthesis: "the combination of ideas into a complex whole"; Synthesize: "combine so as to form a more complex product" [9]. ≈ jossi ≈ (talk) 23:59, 19 March 2007 (UTC)

I've lost track of the argument here. A synthesis is a putting together. We're not allowed to put published facts together in a way that serves to advance a position, unless we can attribute the putting together and the new position to a published source. That's all we're saying. Why is that problematic? SlimVirgin ^(talk) 00:00, 20 March 2007 (UTC)

Because the policy only mentions the word "synthesis" and does not mention that the synthesis must "serve to advance a position." That language was specifically removed from the policy, and attempts to re-introduce it have been met with the usual "please don't change policy" reverts. If you understand what synthesis actually means then I am confused as to how you don't understand that the current policy wording is distorted. -- Y|yukichigai (_ramble argue ^check) 05:14, 20 March 2007 (UTC)

Yes, it does, it says both. SlimVirgin ^(talk) 05:33, 20 March 2007 (UTC)

Well crap. Color me embarassed. Looks like JulesH put the language back in while I wasn't looking. Now I understand your confusion. Sorry. -- Y|yukichigai (_ramble argue ^check) 05:27, 20 March 2007 (UTC)

Also Merriam-Webster: Synthesis: "the composition or combination of parts or elements so as to form a whole.; Synthesize: "to combine or produce by synthesis". ≈ jossi ≈ (talk) 00:02, 20 March 2007 (UTC)

• Meta-comment: Yukichigai, I've been holding off comment on this, because I thought maybe it was just a slip-up, and then two slip-ups, and then three, but this is getting to be a pattern. You really, really need to moderate your commentary on others' alleged abilities to read and understand, as you are well across the line of WP:CIVILity at this point, and that last one borders on transgressing WP:NPA. Loudness and anger don't make an argument stronger. Please also see WP:APR, WP:TEA and WP:MASTODON for further guidance. That said, I agree with you that the term is too subjectively interpretable in conflicting ways for continued use in the policy. I don't think I agree 100% with your linguistic analysis here, though there actually is some evidence that the two terms are diverging in fine-grained meaning, weird as that may be, since one obviously descends from the other. It's a moot point, though, and needn't be fought over any longer - just the fact of the ambiguity is enough to doom this term here as too easy to POV-interpret and fight about. — SMcCandlish [talk] [contrib] ツ 00:51, 20 March 2007 (UTC)
  • Yeah, sorry, I was getting a leeeeeeetle bit frustrated. This is not the first time I've attempted to explain the problem with using the word "synthesis" by itself, and no matter how many times or ways I attempt to explain the concept nobody seems to understand or, more to the point, even pay attention to me. Yelling and being rude, while incivil, at least gets people's attention. -- Y|yukichigai (_ramble argue ^check) 05:20, 20 March 2007 (UTC)

Where to discuss massive dubious sources?

Is there a forum where some sources may be discussed for reliability? In the context of a single article, it may be decided within this article. But what is a source is quoted in a large number of articles for various topics?

My question is triggered by the following issue I sumbled upon Talk:Qazakh, where credibility of a certain "Andrew Anderson" was heavily questioned. It seems that this professor does not have a single peer-reviewed publication (at least his defenders provided none), but it turns out that wikipedia is filled with references to his self-published website. So instead of leading the same debate in 40+ pages, I would like to resolve the issue in one page.

Please comment, both on the particluar case and the applicable forum. `'mikka 01:15, 20 March 2007 (UTC)

Quite realistically, the appropriate forum probably is the article's talk page, with the specific context of what the source is being used for. A given source is not necessarily always reliable or always not, it may be very reliable for certain areas and not so reliable outside its area of focus or expertise. Of course, if it's decided that a source is not reliable in a certain area it's been used for, that discussion could certainly be noted when discussing other articles that use it in a similar vein. Of course, there are some baselines (must be fact-checked/editorially controlled/peer-reviewed in some way, self-published sources should be used with great caution, etc.), that apply across the board. Seraphimblade ^{Talk to me} 05:22, 20 March 2007 (UTC)

Random idea - Attribution (Fiction)

Lately, I've been on the bad end of violating WP:ATT. The problem? I've been writing about a fictional source. I'm not saying that there should be a WP:ATTFIC right now (although it would certainly solve quite a few problems for me, heh), but something could be said about making an official policy that allows, say, the fictional topic's source material to be used without some people getting all uppity about it. Hell, at least one featured article, supposedly one of the best articles produced by the Wikipedia community, uses its source material for almost all of its citations. I already can expect how this article is going to go, but feel free to discuss. Scumbag 02:54, 20 March 2007 (UTC)

Eek, I didnt notice the 'There are no policy innovations suggested' line, so I'm not sure if I'm in the wrong for suggesting this here. Sorry if I am. Scumbag 02:55, 20 March 2007 (UTC)

The fiction work itself can be used as a primary source for information within that source, however, any source of interpretation or synthesis of that source would need its own reliable source. ^High_InBC^{(Need help? Ask me)} 02:57, 20 March 2007 (UTC)

Yeah, but the problem is that where I see using a primary source for information, other people are seeing interpretation or synthesis. Never mind that the information is literally "Look at this. See this? This is what is being said." Sigh. Scumbag 03:01, 20 March 2007 (UTC)

It's a lot easier to step over the line into synthesis then you might think, when using only a primary source. "Darth Vader kills Emperor Palpatine in order to save his son, Luke." That may seem trivial and obvious, but "just from watching the movie", it is synthesis-you don't know what was going through Vader's head. (Nor just from the original movies is it even certain that Vader is Luke's father-he says so, and Luke and Leia believe him, but that's not actually confirmed until later additions to the canon storyline.) On the other hand, of course, "Star Wars Magazine states that Vader killed Palpatine to save his son, Luke" is perfectly alright-it's attributed to someone besides "The editor who wrote this here interprets it so". That's why we should always work mainly from secondary sources, not primary. It's too easy to step into synthesis, and what can be said without it is dry and obvious. Seraphimblade ^{Talk to me} 04:42, 20 March 2007 (UTC)

Quote. Use quotes in the article. Use larger quotes in the ref/notes section to show context. WAS 4.250 04:50, 20 March 2007 (UTC)

An NP-complete problem (not evident information, but easily verifiable)

I have something I wish to add to an article that I discovered on my own. The crux of the problem is that while the information is rather nontrivial, it is very easy to verify on one's own. For those of you with an interest in computer science out there, it's analogous to an NP-complete problem.

Specifically, I'm thinking about adding something to the article on Simlish. When a sim leaves their house, they often say "dag dag". I noticed a while back that if you perform a ROT2 on "bye bye", you end up with "dag dag". Such an observation would never reach a reliable source because it's too isolated and, after all, when was the last time you read a peer-reviewed article on Simlish? Nevertheless, once you see that the two phrases are so simply related, it seems quite obvious that the developers had that relation in mind. So which side to the argument wins out? Is it unconditionally shot down as original research, or is it instead so trivial a fact that it doesn't even need citation because anyone can instantly verify it? I'm more interested in learning how the NOR policy applies here than getting my observation "published". Thanks.--134.173.200.107 06:53, 20 March 2007 (UTC)

That kind of thing is definitely original research, and may not even be true; see what Arthur C. Clarke said about the naming of HAL 9000:

"I've been trying for years to stamp out the legend that HAL was derived from IBM by the transmission of one letter. But, in fact, as I've said, in the book, HAL stands for Heuristic Algorithmic, H-A-L."

--NE2 07:19, 20 March 2007 (UTC)

This merger is a really bad idea

Merging "Wikipedia:Verifiability with Wikipedia:No original research, while also streamlining Wikipedia:Reliable sources into a simpler FAQ at WP:ATT/FAQ" was a big mistake, and the result is completely incoherent. This needs to be reverted immdiately because it is just wrong. It is bad policy all around, and this is not the right way to make policy.

Verifiability and No Original Research are conceptually distinct: they are different things, not the same things. Reliable sources, too, is quite different from the other two, although arguably a subset of Verifiability. The resulting confusion is apparent in many arguments around Wikipedia, as editors are getting confused about two very different concepts.

There is no logical reason for the merge, as the information contained in this unified policy can be more sensibly separated back out into the two separate policies.

As a first step, I am removing the claim on this page that is supercedes the other two, and restoring the other two to their rightful place in the pantheon of Wikipedia.--Jimbo Wales 15:27, 20 March 2007 (UTC)