Jump to content

Wikipedia talk:Attribution/Archive 9

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 5Archive 7Archive 8Archive 9Archive 10Archive 11Archive 15

Problem with using the term "attribution" in the context of this policy

What is not original research

The subsection What is not original research ends with this sentence:

Deductions of this nature should not be made if they serve to advance a position, or if they are based on source material published about a topic other than the one at hand.

I don't see the value of either part of this sentence. The first part seems to be a reference to NPOV, which this policy page is not supposed to cover. The second part I don't really understand at all and I think people trying to apply the policy won't understand it either. Can someone argue for keeping this sentence? --Zerotalk 23:59, 25 January 2007 (UTC)

A key part of the existing policy of WP:NOR that I've often seen missed is that interpretations made to advance a position can be original research. In my mind, one of the more flagrant was Timeline skew theories for The West Wing, which took 3 AFDs (no consensus, no consensus, delete) to get rid of. That was by the second AFD full of citations that showed in show X incident Y was said to happen on Z date, but then built up mounds of argument about how all of these facts were supposed to demonstrate a skew of the timeline from reality, and cited nothing reliable on the timeline of the show as a whole. I've seen this piece of the policy missed so many times, that drumming it home by repetition is in my mind a good thing.
The current part of WP:NOR says "A and B, therefore C" is acceptable only if a reliable source has published this argument in relation to the topic of the article. This language is repeated in the paragraph immediately above the one you mention, and I don't have a reason to want it in both places. GRBerry

without specialist knowledge (revisited)

We have several sentences like this: "It should be possible for any reader without specialist knowledge to understand the deductions" and "Edits that rely on primary sources should only make descriptive claims that can be checked by anyone without specialist knowledge." I know this has been discussed before, but I think more discussion is needed because this rule is really really bad. I'll focus on just one serious problem that in my opinion kills it dead: This rule eliminates almost all journal literature in the sciences. If someone summarises material from the article "Asymptotics and bioavailability in a 17-compartment pharmacokinetic model with enterohepatic circulation and remetabolization" (Horkovics-Kovats and Zlatos, 2006), then I won't be able to check it because I don't have the required specialist knowledge. Does that mean I can revert the edit? Just because I am ignorant of pharmacokinetics (whatever that is)? On the contrary, we should welcome people who are able to read specialised sources and summarise them in plain language in Wikipedia. --Zerotalk 00:32, 26 January 2007 (UTC)

A scientific journal isn't a primary source; it is a secondary source, because it has been peer reviewed. JulesH 07:21, 26 January 2007 (UTC)
Peer review doesn't change a primary source into a secondary source. Everything in a scientific paper is written by the people who created the knowledge. This can vary between disciplines; an extreme example is a mathematical paper which contains facts that came directly from the mathematician's brain. But in any case the first of the two examples I gave is not written in relation to primary sources, but rather all sources. --Zerotalk 08:54, 26 January 2007 (UTC)
The distinction between primary and secondary sources is most useful for history, and becomes less useful the farther you get from that field. The article on Primary source even start with the words "In historical scholarship..." When did the link to that article get removed from this policy?
I tend to consider math research papers to be somewhere between primary and secondary sources. I could see how a paper describing a chemistry experiment, but giving no analysis, could be viewed as a primary source. But any decent math paper not only presents new ideas but also analyzes them and places them in context; that sounds like a secondary source to me. Being peer-reviwed is mentioned by the secondary source article as a characteristic of secondary sources. An advanced book in mathematics may include previously unpublished theorems as well as giving a detailed development and analysis of previously known material. Again, that is a mixture of primary and secondary components. CMummert · talk 13:34, 26 January 2007 (UTC)
Historians, scientists and legal scholars use the term "primary" in different and incompatible ways, largely because they are interested in different things. Wikipedia's usage is that of the historians, so it should be no surprise that it breaks down when used for other fields. Robert A.West (Talk) 16:19, 27 January 2007 (UTC)
As for Zero's original point, the sentences in question all concern deductions made by a Wikipedian, not a deduction of an expert publishing in a reliable source. If a peer-reviewed article on pharmokinetics claims that X1, X2, ..., X17, therefore Y, we may quote or summarize to that effect. Since Wikipedia is a general encyclopedia, the idea as summarized must be reasonably clear to educated readers who are not specialists in the relevant field. In all likelihood, those same non-specialists will be able (with effort) to verify that the article summarizes the article fairly, even if they don't follow the details. What must not be done is for a Wikipedian, based on his own knowledge of pharmokinetics, to say that if Y is true, then Z must also be true, unless the deduction is obvious even to non-specialists. Thus, if statement Y is "A is faster than B", then a Wikipedian may express this as "B is slower than A" without committing original research. (Yes, we have had idiots who argued that such a statement was OR.) Much beyond that requires a source. Robert A.West (Talk) 16:43, 27 January 2007 (UTC)
Yes, but even such elementary uncontentious deductions might require specialist knowledge just to understand the words used in the source. To continue the mathematics example, I think it is absurd to require than someone who never studied any mathematics should be able to look at a research paper and check that it confirms the summary that someone added to Wikipedia. If the source says "the antiderivative of cos(x) is sin(x)" then we should be able to write "the derivative of sin(x) is cos(x)" - this deduction is every bit as elementary and uncontestable as your faster/slower example but specialist knowledge is required to know that. What I'm thinking is that "It should be possible for any reader without specialist knowledge to understand the deductions" should be something like "It should be possible for any reader with the background required to read the source to understand the deductions". That wording probably needs work yet. Another example is when the source is in a rare language; we should not require that someone who doesn't know that language be able to check it. --Zerotalk 22:51, 27 January 2007 (UTC)
Saying "reader", as opposed to "Wikipedian", implies that one is an appropriate reader for the article. IIRC, that word choice was deliberate: we aren't limiting verifiability to the skills of a five-year-old. The article on pharmokinetics requires a certain amount of background to understand. Obviously, we needn't be concerned about attempts to verify by someone who lacks that necessary background. As for elementary calculus, I think it a tremendous stretch to call a course offered in high schools "specialist knowledge". The issue of whether to permit untranslated sources is controversial, and I don't think we have tried to settle that here. My personal opinion is that an untranslated source in French might be acceptable, but one in Hittite wouldn't be. Robert A.West (Talk) 23:19, 27 January 2007 (UTC)

(Evil) straw poll

Do you believe the {{rejected}} tag should be placed on the article? Keep comments short, please.

Support

Oppose

Comment

  • I'm undecided. I was concerned about the direction the editing to this page took, and so I felt it was, for the time being, a failed policy, though I think the group of editors at the very beginning did a great job, and an early version could have worked. If it's to be revived now, I'd prefer to see it as a summary of NOR and V, without any policy change. It was the attempt to add a change that failed, because some people were opposed to any change, and others weren't prepared to settle for a small relaxation but kept wanting to expand the loophole. It had reached the point where some versions undermined the idea of having any sourcing policy at all. SlimVirgin (talk) 17:48, 2 February 2007 (UTC)
The simplicity of using the exact RS and V wording for the exceptions has (hopefully) rendered this largely moot. When the page is live people will be conservative in expanding or contracting anything, so let's take it live with no innovations. Slim is there anything else on the page that you'd consider a policy change? Marskell 17:54, 2 February 2007 (UTC)
  • I have not participated in this page to date, although I supported the initial concept of combining V, RS, NOR into a single unified and coherent policy. I would agree with the above comments, that the way to go about getting this policy accepted is to maintain an exact copy of the existing component-level counterparts, except in cases where there is a clear conflict, where a reasonable compromise must be adopted. I would not deviate from the existing policies until this combo version is approved. Then it can evolve like any other policy. Are there people actually opposing the combined version on principle that V, RS and NOR should remain separate? Crum375 00:48, 5 February 2007 (UTC)
  • I haven't participated much here, but I'm very much in favor of combining V and OR into a single policy, and I think that maintaining the wording of the existing policies is the best way to get ATT accepted. --Akhilleus (talk) 06:27, 8 February 2007 (UTC)

Discussion

Er, guys, we really do not vote on policy proposals (heck, the {{proposed}} tag even says so). If you don't like what the page says, discuss it; that's what people've been doing for the past months. This case is hardly "closed" given the activity here on the talk page. >Radiant< 13:31, 2 February 2007 (UTC)

  • It wasn't a vote. It was a discussion starter. I think we are all grown ups here, and the straw poll was off to a perfectly agreeable atmosphere. Straw polls have long had legitimacy on WIkipedia, you've set a few up yourself in your time Radiant! This is a rather simple poll to resolve a rather simple issue, and saves tag warring. I dread to think that Wikipedia has descended to poll warring now. Whilst m:Polling is evil, this poll doesn't come anywhere near the reasons why that essay exists. There really is no alternative to asking people if they feel the page should be tagged as rejected. It's a binary issue, and thus perfectly formed for polling. Let's not arbitrarily close all polls, because Wikipedia is not a bureaucracy. It's not my poll so I won't reinstate it, but I will assert Marskell's right to do so and respect any decision Marskell makes in regard to this poll, barring a misreading of any potential outcome. ;) Hiding Talk 13:49, 2 February 2007 (UTC)
Thanks Hiding. Absolutely this wasn't a vote on the proposal (hence I chose the rejected tag, rather than the policy tag—obviously no straw poll is going to settle a fundamental policy decision). Discussion is "what people've been doing for the past months". Yes, and it's getting pretty awful. This user would like to know whether it's worth putting time into this or if it's better to move on to other things. I know a number of users (including the person who started the page) think ATT is already dead. Let's see what comes up. Marskell 13:57, 2 February 2007 (UTC)
Put me in the "evil" support category - I think it's so dead that I don't keep up with it as regularly as I used to. It got off to a commendable start, but became sidetracked and abandoned, and since so few editors follow the page now, it no longer reflects any sort of broad consensus. SandyGeorgia (Talk) 14:39, 2 February 2007 (UTC)
I thought this proposal had died a long time ago. I wouldn't normally see any reason to mark something as rejected if it is still being worked on. However, the fact that after three months no consensus has been reached suggests to me that continuing to try and shape this into something acceptable to the community is likely to be a doomed effort, so I'll have to come down as weakly supporting adding the rejected tag (just out of a feeling that there could be more productive things to spend time on). On the actual proposal itself: the third exemption looks just as bad to me now as it did three months ago when the workshop fizzled out. Yomanganitalk 15:10, 2 February 2007 (UTC)
A lot of people think that the third exception is necessary. It allows us to have articles covering topics that we would otherwise struggle to find adequate sources for. Whether these articles should be covered is not the subject of this proposal or of the policies it would replace; the guideline WP:NOTABILITY is the place for such concerns. I've yet to hear a good reason why it shouldn't be included, but I'd be happy to listen to one. JulesH 15:33, 2 February 2007 (UTC)
Actually, this and the policies this page may replace are the places. Notability is a guideline page. Mark me as supporting the removal of the third exemption if that really is the sticking point. The second exemption should really be flexible enough to allow anything the third exemption is aiming to catch. What else do we need to move on to take this forward? Hiding Talk 16:03, 2 February 2007 (UTC)
I think WP:NOT is the relevant policy. JulesH 16:38, 2 February 2007 (UTC)
Per WP:NOT: "Wikipedia is not a publisher of original thought" One of the policies this is intended to replace? WP:NOR. You've proved my point. Hiding Talk 17:05, 2 February 2007 (UTC)

Is ATT intended to replace other existing policies or be an additional new one? I just want to make sure I understand. --Milo H Minderbinder 17:08, 2 February 2007 (UTC)

Not only V and NOR, but RS as well. ≈ jossi ≈ (talk) 17:25, 2 February 2007 (UTC)

Where professional sources offer shallow coverage

Can someone cite an example of a location where professional sources offer shallow coverage that would make self-published sources acceptable? Hipocrite - «Talk» 16:32, 2 February 2007 (UTC)

  • Another example that has been discussed is the use of Usenet sources in documenting the history of usenet. While there is at least one published book on this subject (and perhaps only one; my searches turn up two, but they appear to be the same book published by two different publishers), it is out of print and difficult to find copies of. However, there are plenty of sources available in archives of usenet material, both primary and secondary, that could be useful in writing such articles. JulesH 16:42, 2 February 2007 (UTC)
    • If this is, in fact, the goal of this third clause, I must respectfully insist that it is unlikey this policy will find consensus in it's current form. Hipocrite - «Talk» 16:59, 2 February 2007 (UTC)
    • Hmmm. I can turn up a few sources that appear to cover the history of USENET. [1] & [2]. Maybe the trouble is that people are trying to cover too much ground in their articles. I'm also struggling to think of an instance where a USENET post is a secondary source. Hiding Talk 17:00, 2 February 2007 (UTC)

I support Hiding's changes to the proposal an am currently reviewing the entire policy proposal with an eye to tentatively supporting it as a better explanation of V and OR. Hipocrite - «Talk» 17:04, 2 February 2007 (UTC)

It should be obvious that I support the removal as well. One other thing to consider: take the verbatim wording from RS and V for exceptions one and two, respectively (that would be the second paragraph here and this section) and everyone agree not to touch it again until and unless ATT goes live. (I'm not saying this is or is not better wording—it just means we won't be arguing over policy innovation.) Thoughts? Marskell 17:10, 2 February 2007 (UTC)
Bloody good idea, I've amended accordingly. Hiding Talk 17:20, 2 February 2007 (UTC)
This is starting to look better. SlimVirgin (talk) 17:41, 2 February 2007 (UTC)

FAQ?

I've added a bit more from the NOR policy just to make sure the concept's nailed down. One concern I have is that we're missing the synthesis example from NOR; we put that on WP:ATT/FAQ to avoid taking up space. That synthesis example is one that a few editors have tried to get rid of because they see it as too strict, but it's in fact a very good example of the kind of subtle slide into OR via synthesis that a lot of editors engage in without realizing it. As the FAQ won't itself be policy, I'm a bit worried that, in making this page live, we've effectively given in to the people who wanted to get rid of the example. Would it make the page too wordy to include it here? It's here on the NOR page. SlimVirgin (talk) 18:42, 2 February 2007 (UTC)

I too think that that synthesis example is excellent and you'll recall my badgering you about A + B = C. I have placed the example in; as with the other issues, I think the logic of repeating extant wording is sound. It only adds a bit and the page is still tight.
And I too am concerned about the (non-)policy status of the FAQ. I think we need to decide whether and what in what form we need it. The NPOV FAQ is policy; Francis rightly pointed that forking the page didn't make sense if the policy tag was lost, but in that case it was long-standing wording (it was critical that the pseudoscience section remained policy, for instance). If this FAQ is not to be policy, I'd suggest not having it at all; it will just muddle things. Remember that part of the point is to streamline the flabbiness of RS. We don't need an FAQ that duplicates it. Marskell 19:33, 2 February 2007 (UTC)
If WP:RS is to continue to exist, I think we can ditch ATT/FAQ. The point of the latter was to replace the former, but some editors are quite taken with RS, so there would be a fight. Maybe we should just forget getting rid of RS, at least for now, and focus on replacing NOR and V with ATT. SlimVirgin (talk) 19:47, 2 February 2007 (UTC)
That would be a good first step, although I think that ATT/FAQ is far superior to the current formulation of WP:RS ≈ jossi ≈ (talk) 19:53, 2 February 2007 (UTC)
I've not looked at ATT/FAQ for awhile, but RS wasn't looking good when I last saw it. The alternative is that we make ATT/FAQ policy along with this page, but it would mean getting it in shape, and we'd have the extra battle of getting rid of RS. One step at a time, perhaps. SlimVirgin (talk) 20:18, 2 February 2007 (UTC)
And thank you for adding the synthesis example. SlimVirgin (talk) 19:48, 2 February 2007 (UTC)
I have misgivings concerning dropping the FAQ, principally that it could lead to a considerable amount of bloat, on the other hand, a fight to gain policy status for an FAQ seems unappetizing. I can see the point of dropping the idea. At this point I am pretty much happy to follow SV’s lead, wherever that takes things. Brimba 20:23, 2 February 2007 (UTC)
I look at the current structure as WP:V being a policy, and WP:RS being a guldeline that helps to understand V, but since it is only a guideline, it's OK to violate it when necessary. I think a parallel structure would be to make WP:ATT a policy and WP:ATT/FAQ a guideline. It would be OK to ignore the FAQ when necessary, and in the case of rarely asked questions. --Gerry Ashton 20:38, 2 February 2007 (UTC)
  • Given this discussion over the status of the FAQ I took the liberty of adding the section on anonymous sources onto the page. That includes adding a new third exception. I figure if we feel "Anonymous sources are not acceptable in Wikipedia" then we have to state that in a policy, not a guideline. Hope that's okay, I know we're trying to avoid page bloat, but I think this is an important point best made here. Hiding Talk 21:06, 2 February 2007 (UTC)

Synthesis example

I'm going to object to this. I've written at length on WP:ATT/FAQ and will summarize here:

1) "Given the Chicago Manual of Style's definition of plagiarism, Jones did not commit it" is not original research. It's a simple logical deduction, and those aren't original research. That doesn't mean that the example is okay to use--but it does mean we're giving the *wrong explanation* for why the example is okay to use. The reason it's not okay to use is that the Chicago Manual of Style isn't a reliable source on Jones. The use of a logical deduction is irrelevant--if the Chicago Manual of Style had *directly* said "Jones isn't a plagiarist", it would be equally bad to use.

2) This example appears to refer to Talk:Alan Dershowitz#The debate with Norman Finkelstein, and misrepresents that case. At least one editor attributes it to "James O. Freedman, the former president of Dartmouth College, the University of Iowa, and the American Academy of Arts and Sciences"; James O. Freedman is the one who drew the conclusion from the Chicago Manual of Style, not us. If so, then the statement *can* be used (providing we accept him as a reliable source).

Ken Arromdee 22:27, 2 February 2007 (UTC)

I just checked Dershowitz-Finkelstein_affair. The statement, attributed to Freedman, *is* used. Ken Arromdee 19:48, 6 February 2007 (UTC)
I believe your discussion is at WT:ATTFAQ#Plagiarism_example, and I don't quite follow why you wouldn't like to call this original synthesis. I don't understand your first point: if the Chicago Manual of Style (special edition) would in fact cite that particular case, then why isn't it a reliable source on the issue. --Merzul 00:19, 3 February 2007 (UTC)
Merzul, we're not changing the policy here; we're just copying material from V and NOR, and summarizing it so we don't have two pages. Therefore, to ask questions about why such-and-such is in the policy, you should go to NOR talk. Otherwise, we'll get too bogged down here again. SlimVirgin (talk) 00:24, 3 February 2007 (UTC)
I don't quite follow why you wouldn't like to call this original synthesis.
The reason I wouldn't like to call this original synthesis, or original research in any way, is that original research exempts simple logical conclusions. If some source says "Smith was born in New York City in 1960", I can write "Smith was born in the USA" without violating the original research policy, even though I've technically drawn a conclusion that is not in the source material.
Likewise, "The CMoS defines plagiarism to require X" and "Dershowitz didn't do X" can be used to form the conclusion "According to the CMOS's definition, Dershowitz is not a plagiarist". The process of drawing the conclusion itself requires only a simple logical deduction, and thus fails to be original research.
Of course, this statement is still unusable for other reasons, but not original research.
why isn't it a reliable source on the issue.
Because the Chicago Manual of Style isn't an authority on the Dershowitz case, so any statements it makes about the case can't be used as sources. It would be just the same if the CMoS simply printed the statement "Dershowitz isn't a plagiarist", and we made no deduction (simple or otherwise) to get it. Ken Arromdee 11:40, 3 February 2007 (UTC)
This is hardly a simple logical deduction; it's a complex analysis of Dershowitz's actions based on a complex source. Jayjg (talk) 22:20, 6 February 2007 (UTC)

False attribution

This runs slightly against the current trend (which I otherwise support) to minimize the changes from current policy. But I think now that we have the far better terminology of attribution, then synthesis can be discussed in terms of false attribution. In the synthesis example, one is falsely attributing a conclusion to the Chicago Manual of Style. The article on false attribution has an example where Peter Doran's data on Antarctic temperatures is similarly synthesized to make case on global warming. The good thing with this approach is that it would cover other attribution fallacies like quoting out of context, which are just as serious as synthesis. --Merzul 00:19, 3 February 2007 (UTC)

Merzul, if we start making changes to policy and changes to key terms, this proposal will fail. SlimVirgin (talk) 00:24, 3 February 2007 (UTC)
Yes, I agree with that, I think if people started to think in terms of attribution instead of verifiability, it would have a greater impact than any changes to policy detail. This was just a speculation that I wanted to put on paper, otherwise I fully support what has been going on here in the last few days. Even if a lot of my own junk has been removed :-) Good work! --Merzul 00:40, 3 February 2007 (UTC)
Thank you. I don't think it is a change to call the policy "attribution" instead of "verifiability," because in fact people already talk about whether things can be attributed, and the word "verifiability" is slightly misused in the policy. To verify something implies that we're confirming it to be true, which of course we're not; we're just seeking to attribute it. SlimVirgin (talk) 00:55, 3 February 2007 (UTC)
Yes, yes, yes, this is precisely why I like this proposal. Attribution is not a change to how verifiability is properly understood on wikipedia, and experienced editors like yourself probably already think in terms of attribution, but for a new editor like me, the term "verifiability" is very strongly related to truth. Anyway, I'm going to sleep very happy :) I fully support the current proposal, or the one before my edit, which was just a minor aesthetic thing. The key point is that this proposal should not be rejected because of some details, the main idea is simply too good. Good night! --Merzul 01:13, 3 February 2007 (UTC)

And something else to add: The synthesis example seems to violate its own rules. "A and B, therefore C" is acceptable only if a reliable source has published this argument in relation to the topic of the article.

Yet it uses a reliable source--a message from Jimbo--which makes the argument *in relation to history*--not in relation to an example resembling this one. Ken Arromdee 19:15, 6 February 2007 (UTC)

Hm. Perhaps we should blank the talk and just leave a big note: everthing used on ATT as proposed, is used now on extant P&Gs. Perhaps the example should go, but we're leaving things in that are longstanding until this is live (this was an idea discussed at length above, in order to move the page forward). Marskell 19:25, 6 February 2007 (UTC)
It is my impression that one of the things we are doing here is to clear up parts of the earlier policies/guidelines which are problems. I feel the Dershowitz example is seriously flawed and should not be in here.
And I'm not convinced that it will be possible to modify Attribution after it becomes a policy. It's very hard to get policies modified; if there's something wrong with it, it should be fixed before it becomes policy. If the example doesn't go now, we may never get rid of it. Ken Arromdee 19:44, 6 February 2007 (UTC)
Well, OK: for five months this page attempted to amend policy. Very recently it was suggested that we simply cut & paste from long-standing examples. I understand there's a flip-side to this coin (that it will be hard to amend it once live) but that's part of the point: a conservative process is the only thing that's likely to get us at least out of the starting gate (merging V and NOR into ATT). Marskell 19:58, 6 February 2007 (UTC)
There's nothing wrong with the example; a complicated deduction is a complicated deduction, and claiming that Dershowitz plagiarized based on your reading of the Chicago Manual of Style is inevitably that. It's not adding up a column of numbers. Jayjg (talk) 22:20, 6 February 2007 (UTC)
It isn't a complicated deduction. It's a *simple* deduction. It's utterly clear whether or not Dershowitz's actions, as described, fit the Chicago Manual of Style's definition; they do not. If the Chicago Manual of Style said that plagiarism was a type of canine, would you claim that deciding whether a wolf meets that definition is a "complicated deduction"?
Now, there are certainly reasons why you might not want to accept the conclusion; you might not, for instance, believe that the description of Dershowitz's actions is accurate, or you might not think that the Chicago Manual of Style's definition is an appropriate one to use. But while those are legitimate reasons to reject the conclusion, they are *other* types of reasons, and should be accurately characterized.
The example is also bad because it implies that the Chicago Manual of Style connection was only made by a Wikipedia editor and that no actual source was found for it. This is less of a problem because someone probably won't realize the connection unless they stumble on the Dershowitz article like I did, but it still isn't really what happened; in fact, a source was used and Dershowitz-Finkelstein_affair#Dershowitz's response still includes it. Ken Arromdee 04:14, 9 February 2007 (UTC)
There are certainly better examples of WP:SYNT, and maybe one could start a guideline on synthesis to just collect different examples. This will be a serious problem though, because there is no real consensus on SYNT, see #Analysis of facts from statistical surveys. Maybe having all the SYNT discussion on a separate page and attempt to reach a common understanding on this issue is a good enough reason to create a WP:Synthesis guideline... In any case, I don't think there is much point in trying to change the example in this proposal without that. --Merzul 04:46, 9 February 2007 (UTC)
Also, I think Jayjg is seriously misreading me.
"claiming that Dershowitz plagiarized based on your reading of the Chicago Manual of Style is..."
But I'm *not* claiming that Dershowitz plagiarized. I'm not even claiming that a conclusion saying so belongs in a Dershowitz article. I *agree* that the conclusion is out of place and should not be used as given--just that it is so for different reasons than people like Jayjg think. You can't just take something that's unusable for a different reason and claim that it's unusable because it's synthesis. Ken Arromdee 17:45, 10 February 2007 (UTC)

Anon sources

(copied from SV's talk page) Currently, where in policy is it stated that anonymous sources aren't allowed? Hiding Talk 23:47, 2 February 2007 (UTC)

Hiding, we can put in a sentence somewhere if you want, but it doesn't need its own section, because it's obvious. If the source has no name (and I mean written by an anon, published by an anon), then we don't know who the source is, and can't judge reliability. Therefore, it clearly fails the criteria. If, on the other hand, Bob Woodward publishes material he says he obtained from an anon source, that doesn't affect us, because our source is Bob Woodward in The Washington Post. SlimVirgin (talk) 23:51, 2 February 2007 (UTC)
It's not obvious. It may be to me and you, but I've been wiki-lawyered on this. Now if we never link to anonymous sources then we've got to say that somewhere in policy. All we have currently is guidance at WP:RS. I like what we have at the FAQ, which is where I copied that from, but it can't be guidance. There needs to be something. If you guys can work a sentence in, great. I tried and couldn't work out where it went. Hiding Talk 23:59, 2 February 2007 (UTC)
At the risk of bogging us down again, if we write something about anon sources, we're going to have to say what we mean by a source, because we'll have to explain something like the Bob Woodward example. Jossi and I disagree about this. For me, a reliable source can be either the writer or the publisher of that person's work. So sometimes the reliability attaches to the writer, and we don't care where he's published, because he's so credible. At other times, the reliability attaches to the publisher, and we don't care who the writer is, or the writer's source, or indeed the source's source. Jossi disagrees, and wants us to regard whatever document we decide is reliable, for whatever reason, as our source; for Jossi, people are never our sources. I can see the merit in both positions, but we're going to have to choose one before any sentence(s) on anon sources can be written. SlimVirgin (talk) 01:05, 3 February 2007 (UTC)
Just a quick one, I think it makes a great difference if distancing language or prose attributions are used. When we use a prose attributions then the more important thing is that the person is reliable, but if we are stating a scientific fact and attributing it with a simple footnote, then this better be from a peer-reviewed publication, not a famous physicists personal homepage. So I think this is very difficult and I don't see how you can set it out in the policy. Hopefully I'll be proven wrong! --Merzul 01:21, 3 February 2007 (UTC)
Yes, and I think for that reason (and others), it's better to be flexible regarding what we call a source. Sometimes it's the person we're trusting. Sometimes it's the publication. SlimVirgin (talk) 22:44, 3 February 2007 (UTC)
Well I agree with you Slim. Is Jossi willing to accept a third opinion? Otherwise, just move forwards. I'll keep trying once it's stable. But realistically then the FAQ needs to be implemented too. Hiding Talk 18:01, 3 February 2007 (UTC)
I'm sure Jossi will accept a third opinion. SlimVirgin (talk) 22:44, 3 February 2007 (UTC)
Sure, if we find a way to explain this better. My concern is that editors will miss this fine distinction, and start asserting reliability of a person's opinion published in an otherwise unreliable source, without having a reliable source that reports on that opinion. If a tabloid or a yellow journalism publication says that notable/reliable person Y says X, would it be acceptable to cite that source? I do not think so. That is my concern. ≈ jossi ≈ (talk) 02:10, 4 February 2007 (UTC)
Per yesterday's compromise that less innovation is more, let's leave Hiding's fuller description off, and bump the debate to the future. I like the idea from another editor to leave the FAQ as a guideline when this starts. We can have a go at it now and try to reduce bloat. Marskell 18:08, 3 February 2007 (UTC)
If that's the way it needs to be, fair enough. I was hoping a quid pro quo might be available, but that's not my call. Hiding Talk 18:30, 3 February 2007 (UTC)
Are you suggesting that I'm up to a clever trick? :) Marskell 18:41, 3 February 2007 (UTC)
Heh. No, this time it's me who is attempting a clever trick. :) Hiding Talk 18:46, 3 February 2007 (UTC)
I'm the only one who's allowed the clever tricks, so please step aside. :-) SlimVirgin (talk) 22:44, 3 February 2007 (UTC)
(stepping aside so that SlimVirgin can show us one of her clever tricks). ≈ jossi ≈ (talk) 02:12, 4 February 2007 (UTC)
"his" Atom 04:04, 4 February 2007 (UTC)
That really was a clever trick then! Good job Jossi got out of the way, Slim might have had his eye out. :) Hiding Talk 17:28, 4 February 2007 (UTC)
Or even her eye out. I won't press for details on what might have caused that. :-) SlimVirgin (talk) 01:24, 5 February 2007 (UTC)

Inline citations

PM, could you say what distinction you're drawing between inline citations and material cited in the text? Sorry if I'm being obtuse. SlimVirgin (talk) 01:24, 5 February 2007 (UTC)

I think what he means are detailed prose attributions like "John argues in his book "The Very Reliable Book" (Reliable Press, 2005) that inline citations aren't always needed, while Gordon completely disagrees in a lengthy essay published in a June 2005 issue of Reliability Magazine." I still think these attributions should be footnoted where other information like ISBN and URLS with "last accessed" etc could be added. --Merzul 01:41, 5 February 2007 (UTC)\
Okay, I see. Most of the time, a full citation would be given in a footnote, Harvard ref or embedded link anyway, and most material like this in texts is converted pretty quickly into a footnote by someone else. SlimVirgin (talk)
That's why I wrote "normally". Footnotes exist because including full citations in text is usually clumsy; but let me continue Merzul's example with
The 2007 edition of "The Very Reliable Book" adds on page 45 that "Gordon has completely convinced me".
Sometimes putting references into text is justified. "Page 45" could be made into a note, but if this exchange is worth keeping (an editorial decision), it should be in text.
Also, some citations are shorter;
  • Chapter XVI of The Decline and Fall of the Roman Empire
  • "Singer and Small wrote two papers on this; in their 1976 paper... [list both in references]" especially in an article which does not use Harvard style.
  • "According to 2 Samuel 11:4, David and Bathsheba..."
  • Saevius Nicanor writes (VIII, 34)...
  • "The phrase is taken from the fourth Eclogue of Virgil [followed by quotation]" (This is precise; the whole Eclogue is about a page long)
  • "For example, Voltaire's Candide is the story of a young man who was taught..."
Having a policy mandate inline citation in one of those three forms has been proposed several times, and rejected. Policy should require what we must achieve; how to do it should be left to guidelines, which are not carved in stone. Septentrionalis PMAnderson 17:34, 5 February 2007 (UTC)

Self-published sources again

We've just had a change back to excluding self-published sources except in articles narrowly about the author of the source. This does, of course, exclude plenty of current uses of self-published sources that are totally legitimate. I don't want to have to use WP:IAR to justify the existence of almost every source on Timecube or most of the sources on scientology. Let's fix this problem now: articles that relate to the author (as Timecube relates to Gene Ray and scientology does to L. Ron Hubbard) are a perfectly acceptable place to use a self-published source. Right? JulesH 21:08, 5 February 2007 (UTC)

  • Does this help:

Kelly, Matt (September 29, 2001). "Kelly's i: Site of the Day". The Daily Mirror. p. 23. YES, the guy's mad - but at least he's not boring. Check out the Time Cube rants. {{cite news}}: Check date values in: |date= (help) Hiding Talk 21:38, 5 February 2007 (UTC)

  • Actually, the real problem with the new wording is that it's ambiguous: Scientology is "about" L. Ron Hubbard almost as much as L. Ron Hubbard is; and that's why Dianetics is useful there. Septentrionalis PMAnderson 21:40, 5 February 2007 (UTC)
    • Just to point out, I don't think this is new wording. It's the established policy and guidance copied verbatim from WP:V and WP:RS. Hiding Talk 21:42, 5 February 2007 (UTC)
      • Yes, but it was altered here because of a strong movement, arguably consensus, to do so. I'm not sure the intended effect of "about", to stop pseudophysics from getting into physics articles, will work; this wording permits "Gene Ray says that this physical phenomenon is explained by the Time Cube<:ref>Time Cube site..." almost as much as the one that was there yesterday. Septentrionalis PMAnderson 21:51, 5 February 2007 (UTC)
        • You're going to have to quote me the exact section that is the problem and explain again the problem. I'd argue that there just plain shouldn't be an article on Timecube for starters, if we have no reliable sources to source the article from. What's the issue at scientology? And why isn't The Guardian cited once in that article? [3], particularly [4]. Hiding Talk 22:30, 5 February 2007 (UTC)
          • IMO: If a crank idea is sufficiently notable to be mentioned in reliable sources -- and especially if it was studied as a sociological phenomenon -- then I think it deserves an article. If a idea is discredited among specialists, but believed by enough non-specialists to have a measurable impact on the world, it deserves an article. If a system of belief has millions of adherants, then the fundamental works of that system are primary sources and should be so handled. Robert A.West (Talk) 01:45, 6 February 2007 (UTC)
            • Let's not fix this now. This is verbatim wording from current policy, so it's not rigged one way or another. When this goes live we can debate until we're blue in the face. Marskell 08:09, 6 February 2007 (UTC)
            • To add: I don't think you need to invoke IAR for Timecube. They're modest, but you have at least two reliable sources for the page, that I see. "Ray spoke at Georgia Tech in April of 2005" should not be cited to his blog, however. Marskell 08:15, 6 February 2007 (UTC)
              • Whoever wants to propose a change should do so on the NOR or V talk pages for now. This page is just a summary of the existing policies, and the point of it is simply to have them on one page instead of two, and to get rid of the ambiguity of "verifiability". SlimVirgin (talk) 19:55, 6 February 2007 (UTC)

Differences from current policy

Ok, so let's see what is different?

Are there any more differences that I haven't noticed? --Merzul 22:46, 6 February 2007 (UTC)

Thanks for this Merzul. We can try to audit the page for these things. Marskell 08:01, 7 February 2007 (UTC)

What is not original research?

This is a new section and I understand why this is needed, but I have been dragged into conflicts with editors on obvious issues and such discussions aren't very helpful. I prefer not to have discussions on what is "obviously right" at all... Could we perhaps add something to stress again that the burden of proof is on the editor who wants to include an obvious deduction to find a source for it. That is, if another editor, who doesn't have a disruptive history, is challenging an "obvious" deduction, then a source is really needed, not just further arguments on the talk page. Of course one should read such explanation, but one should have the right to just say "Look, I read you arguments, and I still disagree, please provide a source for you interpretations." I hope you know what I mean... --Merzul 22:46, 6 February 2007 (UTC)

Just to note that I wrote this after auditing the differences and I personally no longer think this is an issue at all. The section "What is not original research?" is very rigid and conservative. Anyway, I think the current version is clear enough about the burden of proof... as the concerns about diligent dissidents shows. --Merzul 16:30, 9 February 2007 (UTC)

Diligent dissident's scythe

  • I've been reading this effort intermittently for months, and have scanned the archives, but haven't yet found a discussion about what has seemed to me the HUGE problem that this policy change seems likely to cause -- not at the policy level, but at the ground level where articles actually get edited: The policy knocks out vast swathes of fairly stable articles built on "common knowledge" (i.e. without citations), or self-published web sources. Fine, these articles will gradually get properly sourced, spurred by this new policy. Yet on an increasing number of articles, a diligent dissident's "sourced" but skewed insertions are held at bay by the consensus of the group. That dissident is being handed a huge scythe here to wreak havoc on otherwise stable articles by tagging or deleting material that is widely accepted but under-sourced, in retaliation for his own "sources" being excluded or minimized. For examples of such dissidence, see Talk:Pretender#Dom Rosario Poidimani, again, Talk:Folke Bernadotte#Neutrality Tag, and Talk:Michael I of Romania#Needs work (in the first and last article, sockpuppets are used, but the reality is that all editors are in agreement except for the dissident). I expect that editors, faced with a dissident newly armed with this policy, will give up rather than try to respond to innumerable [citation needed] tags deployed by the dissident to wear down vigilance on the article and resistance to his edits (see Paul Ilyinsky for an example of how such tags have virtually frozen substantive input to an article since November '06). How is it expected for this to be handled? Lethiere 17:55, 7 February 2007 (UTC)
  • Related to the "persistent dissident's scythe" problem, it seems to me has been the reliance upon well-respected websites for information in obscure or largely abandoned historical fields, where the author has put effort into research and taken advantage of web technology to put it online, rather than into getting their work published in traditional print media. I'm talking especially about research areas where there is virtually no other updated information available, or available in English. Examples include Leo van de Pas Genealogics on obscure royal and noble families which is heavily vetted on soc.genealogy.medieval, and Christopher Buyers' Royal Ark on non-Western dynasties, and Paul Theroff's Online Gotha which updates changes in reigning and deposed Western dynasties and noble families. All three of these sites are constantly in use and under discussion at alt.talk.royalty and other forums that debate royalty & nobility. They are widely cited in Wiki articles on royalty & nobility, because there are simply no published substitutes for these sources, and because the editors are well-known online for accuracy in research, bias-control, and receptiveness to corrections. Lethiere 17:55, 7 February 2007 (UTC)
I would disagree. WP:ATT is based in the already existing policies of WP:NOR and WP:V. On the basis of these policies alone, the articles that you refer to could be "attacked" by dissidents with as much power with, or without WP:ATT. ≈ jossi ≈ (talk) 18:11, 7 February 2007 (UTC)
Agreed, WP:ATT isn't changing the official policy, but it will change the norm -- indeed, isn't that what's intended? Now, challenges are frequently made to content perceived as POV, but much less frequently to content that is undersourced yet widely considered accurate. WP:ATT encourages editors to challenge undersourced material even if it is believed to be accurate. Hopefully, most editors will apply the new policy prudently. But what thought has been given to the impact of editors who see the new policy as a means of putting disputed insertions on the same footing as content broadly deemed acceptable (even if still awaiting appropriate sourcing)? Is it your assumption that this outcome won't result? Lethiere
That is an interesting reading the proposed policy. Which part, exactly, encourages challenges? CMummert · talk 04:50, 8 February 2007 (UTC)
"The burden of evidence lies with the editor wishing to add or retain the material." and "Any unsourced material may be removed..." 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. Lethiere 05:38, 8 February 2007 (UTC)
I really don't see how this policy encourages anything that the old policy didn't already encourage. The section on How to cite and request a source seems far from encouraging the addition of lots of fact tags. In practice, I think most people never read past the first paragraphs, which are quite similar to WP:V. CMummert · talk 12:34, 8 February 2007 (UTC)
We should be intolerant of undersourced material. "Common knowledge" writing was the norm until a couple of years ago, but is now rightly deprecated. Marskell 13:36, 8 February 2007 (UTC)
Exactly. And isn't promulgation & promotion of that policy the point of this project? So given that, how would you deal with a dissident's tags at, e.g., Paul Ilyinsky, inserted because his POV/original research edits were being resisted here & elsewhere by other editors? Please note the kinds of material being challenged, in light of WP:ATT's The burden of evidence lies with the editor wishing to add or retain the material policy. Lethiere 18:47, 9 February 2007 (UTC)
(unindent) It is not the purpose of WP:ATT to tighten any requirements from WP:V. WP:ATT is just meant to be a restatement of the current consensus that replaces WP:V, WP:RS, and WP:OR.
As for the Paul Ilyinsky article, those fact tags are clearly making a point. I would just replace them with an unreferenced template on each section. Acutally I did it: old version new version diff. CMummert · talk 19:36, 9 February 2007 (UTC)


The difficulty in discussing these issues in the abstract is that there is no general agreement what "undersourced" means, as I am sure you are aware. In any case, WP:ATT matches what I believe is that the current (undocumented) consensus on the role of attribution, which is why I am willing to support it in its current version. CMummert · talk 14:11, 8 February 2007 (UTC)
The current policy will replace WP:V and therefore it must re-iterate much of what it says. If you look at the above #Differences from current policy, you will see that the current proposal has slightly reformulated the "burden of proof" concept precisely to discourage editors to challenge undersourced material if it is believed to be accurate. --Merzul 16:31, 8 February 2007 (UTC)

That dissident is being handed a huge scythe here to wreak havoc on otherwise stable articles by tagging or deleting material that is widely accepted but under-sourced, in retaliation for his own "sources" being excluded or minimized. Belated comment: very much agreed. This is under discussion elsewhere in relation to science/medical topics, where it's fairly common to have daft but well-sourced pseudoscientific assertions making notability, but no sourced refutation. This is happening in Mucoid plaque, where one editor has more or less has free rein to post whatever assertion is available for this theory, but very few medical refutations are available. Tearlach 11:54, 14 February 2007 (UTC)

Well, if he refuses too much criticism then how will he establish notability? And if he can establish notability, then we just hope there will be criticism. I think the dilemma works, such people will either have to accept independent reports or they face deletion. In the case of this mucoid plaque, it is highly dubious, and has no media coverage. --Merzul 17:22, 14 February 2007 (UTC)

Intentions of restructure...

I did a little restructure with the intention of easing the transition to this new policy, meaning I want to make it clear this isnt' anything very new, but one section of NOR and one section of V, and attribution is simply the uniting concept. It also makes sense for WP:NOR shortcuts to redirect to the NOR section in the future, and similarly for WP:V except the title might be confusing. What is contained in our section "Reliable sources" is precisely WP:V, and the shortcut WP:RS should probably lead to the FAQ :) At some point the old shortcuts could be deprecated, but it makes sense to have them during the transition. --Merzul 20:08, 7 February 2007 (UTC)

In response to SlimVirgin's revert edit comment: I partially agree, and I was myself concerned with that. I will leave it be for now, except I'm sort of guilty of a similar re-ordering already, and let's at least amend that previous one and move the BLP back up. --Merzul 20:27, 7 February 2007 (UTC)
Except, I can't think of where to put it without disrupting the flow... I'll leave it up to you guys. I short argument about me moving relevant stuff down, note that WP:V is structured like that: it first states you need to cite sources, then tells you what to you should cite from. It's not that illogical when you think about it :) --Merzul 20:31, 7 February 2007 (UTC)