Jump to content

Wikipedia talk:No original research/Archive 40

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 35Archive 38Archive 39Archive 40Archive 41Archive 42Archive 45

research

all research is origional research. it has to start somewhere. —Preceding unsigned comment added by 75.177.141.238 (talk) 23:38, 19 December 2008 (UTC)

Read the policy... it explains what we mean. Blueboar (talk) 02:46, 20 December 2008 (UTC)

RNA magazine to REQUIRE original research posted to Wikipedia

That's right, folks. Original research on Wikipedia. I'm surprised they're not requiring it on Citizendium ;). See Publish in Wikipedia or Perish in Nature about RNA Biology. --TIB (talk) 01:06, 20 December 2008 (UTC)

No, they're not. Note the sentence "The journal will then peer review the page before publishing it in Wikipedia." That is, as the journal accepts articles, it also updates Wikipedia. That is not original research. Phil Sandifer (talk) 01:16, 20 December 2008 (UTC)
They also REQUIRE references if the material is to be in article space. I'm wondering if they are doing this for altruistic reasons or if they are trying to game Google. davidwr/(talk)/(contribs)/(e-mail) 01:28, 20 December 2008 (UTC)
Does it matter? They'll fail fairly miserably, given that we nofollow our internal links. Phil Sandifer (talk) 01:30, 20 December 2008 (UTC)

Directly quoting primary sources

Reading the policy page, it is quite clear that secondary and tertiary sources should be relied upon. What is not discussed in as much detail is legitimate situations under which it is permissible to use, and maybe even directly quote, primary sources. Do such situations exist?

I tend to dislike direct quotes in most cases because they can so easily be cherry-picked and used to enforce a certain POV held by the person quoting. Recently in discussion with some other editors, the argument was brought up that if multiple secondary sources directly quote a primary source, then it would be a permissible circumstance to make a direct quotation.

I believe that the position of this argument in a nutshell is that "if a primary source (and by extension, opinion) is referenced by enough secondary sources, then the primary source itself (opinion) is also notable". I take issue with this argument by stating that verifiability does not always ensure reliability. Furthermore, the fact that a secondary source quoted a primary source to establish their own critique of a subject does not magically make the primary source more reliable. What thoughts do others have here? Spidern 04:09, 23 December 2008 (UTC)

From WP:PSTS: "Any interpretation of primary source material requires a reliable secondary source for that interpretation. Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge." This defines the range of permission to use "primary sources". The footnotes have links to several sources that illustrate some examples, to which I would also add this University of Maryland library guide, which lays out a useful set of examples of primary, secondary and tertiary sources.
..... The issue of reliability is a separate question, discussed as policy in WP:Verifiability#Reliable_sources and as a guideline in WP:RS. Reliability can't be universally determined by asking the question "Is it primary, secondary or tertiary?". ... Kenosis (talk) 04:32, 23 December 2008 (UTC)
What I'm trying to determine is whether it is correct to justify usage of a direct quote by saying that it is well-sourced by secondary sources. I don't believe that it always is. Spidern 05:15, 23 December 2008 (UTC)
As I mentioned, this is a reliability analysis, not a question of whether it's original research. This is particularly the case with a direct quote, where no arguments would properly arise over whether your words accurately express the content of the source you're proposing to use. There's no need to justify its use-- if you reasonably think its reliable, use it without apology. If someone else catches some set of factors indicating it may not be reliable (the nature of such factors will depend on what kind of topic it is, of course), then in due course it may get pointed out, which would generally trigger a more thorough WP:RS analysis. (NB: Many experienced contributors to WP would, I think, appreciate your diligent caution. WP:BE BOLD exists as a counterbalance to excessive cautiousness-- that feeling of wanting to get it exactly right the first time. No need to be a perfectionist-- if it's all that controversial, at some point one or more other users will likely respond accordingly.) ... Kenosis (talk) 16:47, 23 December 2008 (UTC)

Inside information

I have recently added this new section based on discussions elsewhere:

Wikipedia is not a crystal ball. Inclusion of "inside information" that one is aware of from his/her employment, studies, or connections, but is yet to be published in any newspaper, web site, or publication, or broadcast on the air, is not permitted on Wikipedia. For example, this may include news on the latest scientific breakthrough, a trade between sports teams, layoffs planned by a company, or legislation about to be passed. Not to mention, this may be a violation of the secrecy policies of one's employer or agent, or even the law.

I feel this mention is necessary because I have seen such additions myself.

Hellno2 (talk) 19:31, 23 December 2008 (UTC)

  • Sorry to disagree. There are numerous types of original research. Ho reason to list them there. "inside info" is hardly a major source. Unreferenced information of any type may be challenged at any time. Adding not critical text unnecessaarily bloats the policy and makes it difficult to read. Mukadderat (talk) 19:37, 23 December 2008 (UTC)
  • No. Unpublished claims are never allowed at wikipedia, period. The policies already have sufficient wording to that effect, and they should stay as short as possible. If you see rumor-type content added to articles, remove it-the policies are already utterly against unpublished sourcing. Professor marginalia (talk) 19:41, 23 December 2008 (UTC)
  • No need to spell this out in the policy... it is already covered from multiple angles by multiple policies and guidelines. WP:V, WP:RS, and WP:NOR all address "rumor" and unpublished information. Blueboar (talk) 19:46, 23 December 2008 (UTC)
  • Blueboar is right. Material that is "yet to be published" is unverifiable, and thus -- by definition -- fails WP:V. -- Fullstop (talk) 21:28, 23 December 2008 (UTC)

Description vs. interpretation

It seemed like, by the end of the last conversation on this, at least everybody largely was willing to grant that the description vs. interpretation wording was unclear. I proposed new wording, which seemed controversial - but I think this is something we really need to shore up, because right now this policy does not provide coherent or useful guidance on this point. Does anyone have any better proposed wording? Phil Sandifer (talk) 20:00, 23 December 2008 (UTC)

Just for the record, I don't find the wording unclear at all, and would be quite happy to leave it as is. That said, I do understand that you (and a few others) do find it unclear and think a rewrite is needed. I will be quite willing to consider any wording you may propose. I will base my agreement or disagreement of any proposed language based upon whether I think it mearly restates what the policy currently says using different language, or whether I think it changes the policy. A restatement in different language I am likely to accept; I am unlikely to accept a change in policy. Blueboar (talk) 20:36, 23 December 2008 (UTC)
Well, to be fair, my problem is that the current language is unclear, so I'm hard-pressed to figure out how to restate it so that is says the exact same thing, given that right now it seems to say nothing. Perhaps you could make an attempt at a clearer explanation of this supposed description/interpretation divide. Phil Sandifer (talk) 22:11, 23 December 2008 (UTC)
I'm not sure how you see it this way, perhaps you are getting a little distracted by the deconstructive process you touch on above, but average readers are not operating on this level and seem to get the point. Semitransgenic (talk) 15:06, 24 December 2008 (UTC)
You're joking, right? Have you looked at fiction articles? The line on what is and is not allowable is a mess over there, with regular edit wars. If there is a clear line, by all means, explain it to me. Phil Sandifer (talk) 15:59, 24 December 2008 (UTC)
sorry that's something I meant to mention, I personally do not think editors should be describing plots, or talking about fictional content, unless they are referring to sources that have already done so, but that's just my personal view. There are enough book reviews, video game reviews, film reviews etc. that can be drawn upon. Also, film, book, video game promotional information, although primary, would be better than nothing. Yes, I have come across many fiction articles that I thought should simply not exist, if the policy infringements are considered, but the problem is endemic, and unless there is a concerted drive to address this it seems like a battle lost. Semitransgenic (talk) 16:13, 24 December 2008 (UTC)
So, actually, your position is that use of primary sources is outright forbidden? Phil Sandifer (talk) 16:25, 24 December 2008 (UTC)
Phil, we have explained it to you... several times. But here it is again: A very basic summary of the surface plot of a work of fiction is assumed to be easily cited to the primary source (the work itself) and thus it is not considered a violation of WP:NOR for an editor to write such a summary. Any analysis, or interpretive statement beyond that would require a secondary source for it not to be OR. Blueboar (talk) 16:25, 24 December 2008 (UTC)
I would personally rather we didn't have editors summarizing primary sources but instead referred to a secondary source that has already undertaken such a task, but this does not mean to say I think primary sources should be forbidden, it depends on the context of their usage. Semitransgenic (talk) 16:40, 24 December 2008 (UTC)
Here's my problem with that explanation - it's still relying on a surface/depth distinction that I think is artifial and non-universal. Superficiality and depth are not inherent features of a work of fiction. How can I tell whether a given statement is surface or not? Phil Sandifer (talk) 19:25, 24 December 2008 (UTC)
I know nothing about this, it's not my area, I can only guess you are talking about the distinction between writings that are very literal and those that are deeply allegorical. In the case of the latter, I assume it would be more appropriate to lean toward secondary sources in discussing the content. You, it seems, are a specialist, if other specialists on wikipedia operating in the same domain share these concerns, make yourselves heard, give clear examples, and give a workable alternative, then gage consensus Semitransgenic (talk) 19:40, 24 December 2008 (UTC)
I would say that if there is any doubt as to whether a given statement qualifies as a legitimate NOR summary of the basic surface plot... assume that it isn't. Go find a secondary source for it. After all, it is never wrong to cite a secondary source to support a statement. Blueboar (talk) 20:57, 24 December 2008 (UTC)
That doesn't really answer the question, though - how does one tell if something is surface or not? That says what to do when in doubt, but that's not what I asked. Phil Sandifer (talk) 01:17, 25 December 2008 (UTC)
The surface plot of a story includes the events that happen to the main characters... where they go and what they do( X character goes to Y place and does Z action) It does not include the characters' motivations for going to the place or doing the action. The surface plot does not include the characters thoughts about what they do. It sticks to "X then goes to Y and does Z". Blueboar (talk) 07:03, 25 December 2008 (UTC)
That's a specific example, but I was looking for the general case - in any given primary source, how do I tell if a given claim about it is surface or not? Phil Sandifer (talk) 14:21, 25 December 2008 (UTC)
If the statement simply discribes where the characters go or what they do in the story, then it is a discription of the surface plot. If the statement goes beyond that it isn't. Simple. Blueboar (talk) 14:43, 25 December 2008 (UTC)
Not all primary sources are works of fiction. The example given is plot summary, but I mean more generally about primary sources. Phil Sandifer (talk) 16:48, 25 December 2008 (UTC)
I have been using "surface" purely as a modifier in conjunction with the word "Plot". Non-fiction does not have a plot to summarize, so the modifier "surface" would not be relevant to non-fiction. For non-fiction the line between appropriate use and inappropriate use (OR) of a primary source is drawn in a different way. It is not OR to give a basic description/paraphrase of what is specifically stated in a primary source. Such basic descriptions are not OR, because you can point (and cite) to the primary source itself, and any reader who looks at the source will be able to see that the source does in fact say what you claim it says. As soon as you start to go beyond what is specifically stated in the source, as soon as you include interpretation or analysis of the source, you need to point to a secondary source that contains this interpretation or analysis. In other words, you have to show that the interpretation or analysis originated with someone other than you, a Wikipedia editor. And again, if there is any doubt as to whether something you want to say is verging on OR or not, you are better off assuming that it is OR, unless you can find and cite a secondary source to support it. Blueboar (talk) 17:23, 25 December 2008 (UTC)
Ulysses and Finnegan's Wake are said to have plots of which an aspect is "on the surface". It makes my head spin. On the other hand, Charlotte's Web, another famous allegorical work, has a surface plot that is fairly straightforward, the basic story of which most reasonable persons who've read it would agree what it is. Maybe it's best to simply apply the rule on a case-by-case basis. The rule, as presently expressed, is: "Primary sources ... may be used in Wikipedia, but only with care, because it is easy to misuse them. Any interpretation of primary source material requires a reliable secondary source for that interpretation. Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge." ... Kenosis (talk) 20:24, 25 December 2008 (UTC)
Kenosis, we got into this discussion because Phil seems to have a problem with the very section of the policy you just quoted... he does not seem to understand the distinction between interpretation and descriptive claims (at least not where articles on fiction are concerned). Perhaps you can do a better job of explaining it to him than I have. Blueboar (talk) 23:21, 25 December 2008 (UTC)
The events which take place in Ulysses are straightforward--the understanding of the connection between them is what needs elucidation--but this is an exceptional example in two respects--first, that the book is exceptionally complicated, second, that multiple reliable descriptions of the events and their connections are readily available as sources. Few books are quite as complicates; few books have the plots as much discussed. the ordinary case is where we have the description of a plot in a secondary or tertiary source. If the source is a reliable textbook , such a even published notes for students, those retellings are usually reliable. If the source is a review, or a summary at a place like IMdB, they are not in my opinion reliable,and would much prefer a direct account from the work, with key events perhaps sourced to particular pages or scenes. Too many of our plot summary pages are made unencyclopedic by paraphrasing such unreliable sources, typically written as teasers. I interpret NOT TV GUIDE as meaning we write real summaries when necessary, rther than use such sources. DGG (talk) 03:15, 26 December 2008 (UTC)
Good point about the events being extremely complicated when the book is taken as a whole, so much so that the editors of Ulysses (novel) apparently chose to reduce the overall plot description to one sentence in the article lede. The plot descriptions given in Ulysses_(novel)#Structure, in my opinion, are a very reasonable application of the policy, since they describe each chapter's plot and structure on face value. Even Ulysses_(novel)#Episode_3.2C_Proteus, which refers to the famous stream of consciousness technique, is a reasonable description that does not appear to transgress the general prohibitions of WP:NOR. As is typical of many WP articles, there could be better application of WP:V via more inline citations, although that basic issue is not necessarily relevant to this discussion here at WT:NOR. ... Kenosis (talk) 23:56, 26 December 2008 (UTC)

To be clear, I'm asking in general, not just for fiction. How, when reading a primary source, do I tell whether a given statement about it is descriptive or interpretive? Phil Sandifer (talk) 17:12, 26 December 2008 (UTC)

These are simple words. (from dictionary.com)
  • Describe: 1) to tell or depict in written or spoken words; give an account of: He described the accident very carefully. 2) to pronounce, as by a designating term, phrase, or the like; label: There are few people who may be described as geniuses. 3) to indicate; be a sign of; denote: Conceit, in many cases, describes a state of serious emotional insecurity. 4) to represent or delineate by a picture or figure.
  • Interpret: 1) to give or provide the meaning of; explain; explicate; elucidate: to interpret the hidden meaning of a parable. 2) to construe or understand in a particular way: to interpret a reply as favorable. 3) to bring out the meaning of (a dramatic work, music, etc.) by performance or execution. 4) to perform or render (a song, role in a play, etc.) according to one's own understanding or sensitivity: The actor interpreted Lear as a weak, pitiful old man.
Does this help? Blueboar (talk) 18:45, 26 December 2008 (UTC)
  • Not really at all - observe the similarity between "explain" and "give an account of," for instance. The most sizable difference that I see is this issue of "hidden meaning," but that brings up a host of problems too - hidden suggests that the meaning is not intended to be found - that it is a secret meaning. So that would not rule out intended subtext. And again, the key question I have - how do I know, reading a primary source, whether a meaning was hidden or not? Phil Sandifer (talk) 23:50, 26 December 2008 (UTC)
When describing fiction as presented in the primary source (the work of fiction itself), a statement is descriptive if it summarizes the fiction on "face value" or makes a statement about some particular portion of it which would be apparent to reasonable persons with a general education upon a plain reading of the primary source. It's interpretive if it explains, explicates, or elucidates the fiction in ways not evident upon a plain reading of the story. If there's disagreement about which is which, it should be discussed or argued among those involved in the applicable WP article. Any disputes should be regarded as a "content" dispute, not a "policy" dispute. As an editorial policy matter, the difference between descriptive and interpretive is fairly straightforward, even with respect to fiction, and even if it doesn't have a hard-and-fast borderline between what's descriptive and what's interpretive. ... Kenosis (talk) 22:48, 26 December 2008 (UTC)

What makes a reading plain, or on face value? What any normally educated reader would be expected to glean from it? In which case, that is, I think, a better explanation than descriptive/interpretive. Also, what do we do with specialist texts? A reasonable person with a general education would get relatively little from, say, a scientific paper, or from a book by Derrida - but then again, such readers aren't really part of the intended audience of such works - so I'm not sure how a general education is relevant to such works. Do we mean a reasonable editor with a general education, or do we mean the sort of reader the work is aimed at? Phil Sandifer (talk) 23:50, 26 December 2008 (UTC)

That's at least a two-part question.
.. RE "What makes a reading plain, or on face value? What any normally educated reader would be expected to glean from it? In which case, that is, I think, a better explanation than descriptive/interpretive." :
.......... This is a reasonable definition that's already given in the policy page w.r.t. use of primary sources. I quoted it above.
.. RE "Also, what do we do with specialist texts? A reasonable person with a general education would get relatively little from, say, a scientific paper, or from a book by Derrida - but then again, such readers aren't really part of the intended audience of such works - so I'm not sure how a general education is relevant to such works."
.......... Yep, that is also dealt with in the present language of the policy about use of primary sources. I'll quote it again. It reads: '"Primary sources ... may be used in Wikipedia, but only with care, because it is easy to misuse them. Any interpretation of primary source material requires a reliable secondary source for that interpretation. Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge." Thus, when dealing with something obscure like Derrida or implications of quantum physics, quote Derrida as may be deemed appropriate by consensus of participants in the article, but if there's any question about what Derrida meant, use the secondary and tertiary sources in support of such interpretation. Same with interpretations of quantum physics and other obscure material that require specialist knowledge. That, essentially, has been the policy for quite some time now-- for a couple years at least. ... Kenosis (talk) 00:06, 27 December 2008 (UTC)
That has not been the methodology followed in practice on pages on specialist topics. In fact, I have trouble thinking of a page on a specialist topic that works that way. The widespread declining to follow this policy makes me doubt its accuracy. Does anyone know the origins of this phrase? I'm really curious if it was proposed originally to forbid the use of difficult sources, or if it was originally intended to preclude specialist readings of general sources. Going back and figuring out what this phrase came from and what it was intended for seems to me important in clarifying it.
I would also suggest that "easily verifiable" is not a very good synonym for "would be expected to glean." But based on this, I would suggest that the description/interpretation distinction (which is misleading at best) should be removed in favor of the reasonable person test, at the least, and that the specialist knowledge phrase should be re-examined, both in terms of what accepted practice actually is, and in terms of what the phrase was actually inserted to accomplish. Phil Sandifer (talk) 00:19, 27 December 2008 (UTC)
It may not be optimally followed, but it is the present policy w.r.t. this basic issue, one which has remained stable for an extended period of time. Similarly, WP:V may not be optimally followed, but it too is the present policy (I leave aside the residual debate about the difference between "verifiable" and "verified with an inline citation" ). The wiki has an extremely broad range of topics for which the core content policies need to account. A great deal of work and discussion went into the above-quoted passage about primary sources. At present, there appears to be little cause to change this aspect of policy, at least lacking a broad wiki-wide consensus to do so. ... Kenosis (talk) 00:39, 27 December 2008 (UTC)
Actually, as I point out in the section above, the specialist knowledge line was inserted with a misleading edit summary, and was never the subject of a prolonged discussion. It seems to me clear that this phrase is an instance of a change made by one person in the days when policy changes were much easier and made much more carelessly, and that it has persisted with no serious thought about its meaning or consequences. To say it has consensus is deeply misleading - that would imply that some discussion has taken place. All there is is an extended failure to actually look at the line. Phil Sandifer (talk) 02:23, 27 December 2008 (UTC)
In fact, the words "without specialist knowledge" were very extensively discussed in the latter part of 2007, as was the entirety of WP:PSTS. It's a key phrase of the policy, so as to help prevent pretense of legitimacy to statements such as the one Professor Marginalia refers to above, made by an anon IP at Talk:Meme#History. It's also a key phrase intended to help prevent further proliferation of quack theories in general, which was the original policy reason for the very existence of WP:NOR. ... Kenosis (talk) 02:40, 27 December 2008 (UTC)
Can you link this discussion? Phil Sandifer (talk) 02:53, 27 December 2008 (UTC)
Maybe try archives 20 through 34, and type "without specialist knowledge" into your "find" box. It's mentioned something like 70 times from September 2007 through January 2008. ... Kenosis (talk) 05:35, 27 December 2008 (UTC)
Also, Phil, one method of dealing with this issue of levels of obscurity has involved splitting topics into basic introductory versions and advanced versions of such topics, intended for general audiences and specialized audiences, respectively. A number of examples of such bifurcation can be found at User talk:Kenosis/Research2. ... Kenosis (talk) 00:47, 27 December 2008 (UTC)
I think that that is a very good approach, but it seems incidental to the question of the use of advanced sources. Phil Sandifer (talk) 02:23, 27 December 2008 (UTC)
Could you give us some examples of pages on specialist topics where this "methodology" has not been followed in practice or where editors have declined to follow the policy? Blueboar (talk) 00:34, 27 December 2008 (UTC)
Jacques Derrida summarizes difficult Derrida works. As does Jacques Lacan. In science topics, Single-molecule magnet relies primarily on scientific journal articles that require specialist knowledge to understand, and that are primary sources as they are the places where the research was first presented. I suggest anyone who supports the idea that sources requiring specialist knowledge cannot be used attempts to cut the "OR" from any of these articles and see how it goes. Phil Sandifer (talk) 02:23, 27 December 2008 (UTC)

Looking over this section, I see three readings of the descriptive/interpretive distinction:

  1. The distinction is meaningless (that would be me)
  2. The distinction allows only things that are explicitly and clearly stated in the text, precluding any reading beyond the direct meaning of the words. (Blueboar)
  3. The distinction allows only things that a reasonable editor would agree the text means. (Kenosis)

I hope I am correct, Kenosis, in reading your agreement with my comment about hidden meanings as meaning that implications that it is clearly intended that the reader will see, and that a reasonable reader would see are allowable even if they are not stated in the text.

I would suggest, however, that if three readers come to substantially different conclusions about the distinction, the language is unclear. I would like to move towards a functional test rather than a definitional one - that is, to focus on the "reasonable person" test rather than the simple definition of interpretive/descriptive. This, I think, makes the standard clear, which it is, apparently, not. Phil Sandifer (talk) 06:55, 27 December 2008 (UTC)

You slightly misinterpret my stance. Let me rephrase it: "In order to discuss things that are not explicitly and clearly stated in the text, or to discuss any reading beyond the direct meaning of the words, you must cite a reliable secondary source for that thing or reading." In other words, we most certainly can include discussion of quite complex ideas and arguments in our articles, as long as these ideas and arguments come from a reliable secondary source. What we can not do is include our own ideas and arguments, as that would be Original Research.
I took it as obvious that we were still talking about the use of primary sources, but yes. In any case, we have three very different views. Actually four, as I'd forgotten Semitransgenic's "Primary sources cannot even be used for plot summary" view of what description/interpretation means. This is clearly a disfunctional part of the policy. Phil Sandifer (talk) 15:15, 27 December 2008 (UTC)
Phil, perhaps this will be helpful. If you look at The Thriving Cult of Greed and Power#Synopsis, you'll note it cites only secondary works. That is how it should be; we do not rely on the views of Wikipedia editors to cull primary sources for the salient and important points, but instead rely on reliable secondary sources to do so. Jayjg (talk) 01:05, 28 December 2008 (UTC)
And where the secondary sources are rich enough, good enough, and based enough on summary to do so, that's lovely. The rest of the time, we get into serious problems. I mean, Derrida is one of the most cited philosophers ever. Any given work of his has dozens of citations and responses. But it often doesn't have a lot of summaries. So we end up losing a rather key aspect of the subject - a description of the subject. Phil Sandifer (talk) 02:14, 28 December 2008 (UTC)
I'm rather doubtful that there aren't any good summaries of as widely discussed a topic as Derrida. In fact, I strongly suspect that there are far more good secondary sources on Derrida than there are on Scientology. Jayjg (talk) 02:59, 28 December 2008 (UTC)
Good? Yes. Lots of very good ones. But summaries? You've got to understand, in literary theory, there's not a big focus on intro texts. Most intro to theory courses don't work from textbooks summarizing theorists. The subject is taught from primary sources. So unlike scientific topics, there's not a big or standard textbook that summarizes the knowledge. There are sources dedicated to summarizing Derrida, but this would be like writing an article on a computer science topic by using Java for Dummies - sure, it's a summary, but it's not the quality we want to use. But Derrida is a tip of the iceberg here - plenty of academics and scholars who's work passes WP:N do not have even that level of summaries. Are we meant to describe these subjects without providing the minimal service of describing their thought and work? Phil Sandifer (talk) 13:29, 28 December 2008 (UTC)

Specialist knowledge

I found the edit that inserted the "specialist knowledge" claim - it was [1]. There was no discussion prior to its insertion, and none of the discussion following (which can be read at the bottom of [2]) dealt with specialist sources. Looking at SlimVirgin's history at the time of the edit, it's clear that the edit was intended to deal with situations like this one: [3]. The case here is use of a source to support claims well beyond what was made in the source - not a case of a source targeted at a specialist audience.

I can find no evidence that this line was ever seriously thought about, considered, or worked through, or that any consensus for its insertion or intentions was ever formulated. It appears to me that it was an ill-considered insertion from the early days when policy formation was easier and done more carelessly, and one that has never been thoroughly examined.

Can anyone show any evidence that a consensus has ever been demonstrated on Wikipedia that sources aimed at a specialist audience cannot be summarized? That is, that this view formed because of active thought on the subject and not because of, to be only mildly ironic about it, interpretive claims about an undiscussed change to the policy three years ago? Phil Sandifer (talk) 02:23, 27 December 2008 (UTC)

Phil, regarding the statement in your second paragraph just above, If you were unable to find evidence of serious consideration of the phrase "without specialist knowledge", that would be because you did not search through the archives covering the time period I stated above, which was "the latter part of 2007". Maybe try archives 20 through 34, and type "without specialist knowledge" into your "find" box. It's mentioned something like 70 times from September 2007 through January 2008. Far from being ill-considered and careless, t's been hashed over and over and over again. Lacking something new here, I'm going to go focus on something else for awhile. ... Kenosis (talk) 05:43, 27 December 2008 (UTC)
I have now read all uses of the word "specialist" in all 14 of those pages. I stand by the fact that there has been minimal discussion of this issue. Furthermore, every time it has come up, it has, frankly, been controversial. In fact, I can find few discussions where this concept is in any way agreed upon. The main argument appears to be "Well if we don't have that rule in place we have no quality control because we don't check expertise." Which is a bizarre claim at best, as it seems to suggest that the available alternative - that articles should be written by people who don't understand the relevant sources - is desirable.
If a person does not understand the sources relevant to editing an article on a given topic, they have no business editing the topic. But on the other hand, it remains an absurdity to suggest that an article like single molecule magnet or Jacques Derrida can be written without going to difficult primary sources. We may as well delete both of those articles if primary sources that require specialist knowledge to understand are forbidden, because the topics are flat-out impossible to write a useful, NPOV analysis of.
Which is something you've agreed with before, it appears. The issue is that these are somehow "exception" articles. Just like, apparently, difficult works of fiction (which are another can of worms - is the issue with Finnegan's Wake that it requires specialist knowledge? It seems like not. It seems like that book isn't specialized, it's just plain hard). What I wonder is if there are not more exceptions than articles where this rule is useful. It appears to be a classic case of a rule designed for the most pathological of bad articles (in fact, we know exactly what article it was designed for - Animal rights. SlimVirgin wrote it specifically to forbid something she found frustrating on that article), with no regard to the general case. And in all of the discussions in archives 20-34, there is precious little attention paid to the general case. Has a discussion ever actually concluded with a decision regarding this? I've seen inconclusive discussions and no discussion, but as it stands, the specialist knowledge aspect of this policy appears to stand primarily on the weight of tradition - it was around for a long time, so it by default became consensus despite the fact that nobody had ever really thought about it. And even though every time it was thought about it was contentious, because it's old, it stays. This despite the fact that numerous articles, including the vast majority of our advanced topics in any academic field, routinely violate this policy in a way that I am unable to believe anyone would have a serious chance at changing were they to go in and start citing this policy.
I do not believe that the "no specialist knowledge" rule has consensus. Unless someone can show a discussion where the issue was actually worked through and the consensus among participants was that this is policy, between its dubious origins (modifying policy to solve a particular editing dispute is not a sound approach) and its routine lack of application, I am strongly inclined to remove it as a part of policy that lacks community consensus. Phil Sandifer (talk) 06:48, 27 December 2008 (UTC)
Phil, I have to object to the idea that we have a "no specialist knowledge" rule. The whole point of this policy is that specialist knowledge is what we want... as long as it is cited. Put the statement about "specialist knowledge" into context an it makes far more sense... "Any interpretation of primary source material requires a reliable secondary source for that interpretation. Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge." This limitation only applies when we are citing a primary source, and only when there is not a secondary source being cited. It does not ban specialist knowledge. It simple says we need to cite secondary sources to support any statements that require specialist knowledge (ie we have to cite the specialists). Blueboar (talk) 15:35, 27 December 2008 (UTC)
I think it's reasonably obvious we're talking about the primary sources section here. You will forgive my saving of keystrokes in not specifying "primary sources" every other sentence. The fact remains that advanced topics in almost any academic field are uncoverable under this rule, as is well known by anyone who edits them. And yet I challenge anyone to eviscerate Deconstruction based on the fact that it is based primarily on summaries of primary sources and get that move to have consensus. Or to show that the language in this policy has ever demonstrated consensus. Phil Sandifer (talk) 18:27, 27 December 2008 (UTC)
the article Deconstruction should be completely rewritten, the lack of secondary sources is hugely problematic. Semitransgenic (talk) 20:34, 27 December 2008 (UTC)
The article Deconstruction is pretty good, and were you to try to slash and burn it you would be swiftly and correctly reverted. If changes to enforce policy have no chance of garnering consensus, it suggests a serious problem. It could use more secondary sources, however the summary of the main ideas of deconstruction is both pretty good and the single most important aspect of the topic. Phil Sandifer (talk) 22:39, 27 December 2008 (UTC)
"The fact remains that advanced topics in almost any academic field are uncoverable under this rule, as is well known by anyone who edits them" - I would agree that it is extremely difficult to cover advanced topics using just the Primary source... or rather it is very difficult to so without slipping into OR. However, this does not mean such topics are uncoverable. In fact, it is fairly easy to cover such topics if you use reliable secondary sources. Blueboar (talk) 21:34, 27 December 2008 (UTC)
Derrida and Deconstruction are poor examples, there are reams of secondary sources available on both topics. I beg to differ, but in both instances it would be possible to avoid using Derrida directly; there is nothing an editor can say on either subject - in the context of writing an encyclopedic article - that will not already have been published in a reliable secondary source and I'm a little suprised Phil seems to believe otherwise. Semitransgenic (talk) 22:33, 27 December 2008 (UTC)
The issue is not whether reams of secondary sources exist, but rather whether an encyclopedic overview can be given using them entirely. Of course the secondary sources have a huge role to play in these articles. However the primary sources also have a very large role to play. That's the issue - that in addition to secondary sources, primary sources are needed to provide an encyclopedic overview. Phil Sandifer (talk) 22:39, 27 December 2008 (UTC)
Phil Sandifer stated:"In science topics, Single-molecule magnet relies primarily on scientific journal articles that require specialist knowledge to understand, and that are primary sources as they are the places where the research was first presented". Academic journals are generally peer reviewed i.e. a body of experts has confirmed that such research adheres to the accepted scientific practices of one discipline or another, I don't see how you can make any comparison between this and what you were originally discussing. Semitransgenic (talk) 15:07, 27 December 2008 (UTC)
Well, that ought to also allow in Derrida's books (all published under university presses), and a host of other specialized works that come via academia. But the policy as it stands makes no such exception. Phil Sandifer (talk) 15:15, 27 December 2008 (UTC)
I'm not sure I get your point, if there are a host of academic works, which have been peer reviewed, there is nothing to prohibit there use. In terms of Derrida as a primary source, there are many secondary sources that deal with his life and his ideas, an editor doesn't need to refer directly to Derrida's work, there simply is no need. Semitransgenic (talk) 15:28, 27 December 2008 (UTC)
First of all, that's not true. One is very hard pressed to cover Derrida's work encyclopedically without going and summarizing the work. (And to attempt to do so creates a grotesque NPOV problem. For one thing, Derrida comprehensively addressed several of his critics - to the point where secondary sources generally did not continue to engage that bit of criticism. So if we work entirely from secondary sources, for instance, John Searle's attack on Derrida goes un-answered - a huge NPOV problem. Second of all, Derrida is published by university presses - the situation is analagous to single molecule magnets. If we can summarize the academic-published work on single molecule magnets in Single molecule magnet, we ought to be able to do the same with Derrida in Jacques Derrida. However, current policy seems to forbid both. Phil Sandifer (talk) 19:08, 27 December 2008 (UTC)

(outdent) with little effort I found Derrida, Searle, Contexts, Games, Riddles, Edmond Wright, Source: New Literary History, Vol. 13, No. 3, Theory: Parodies, Puzzles, Paradigms (Spring, 1982), pp. 463-477. This is a a secondary source that has already addressed the issue you raise. There is also Taking It Personally: Reading Derrida's Responses, Reed Way Dasenbrock, College English, Vol. 56, No. 3 (Mar., 1994), pp. 261-279. Published by: National Council of Teachers of English

For the record, having looked at the Derrida article someone is already taking libertties with OR, for example:

Derrida began speaking and writing publicly at a time when the French intellectual scene was experiencing an increasing rift between what could broadly speaking be called "phenomenological" and "structural" approaches to understanding individual and collective life. For those with a more phenomenological bent, the goal was to understand experience by comprehending and describing its genesis, the process of its emergence from an origin or event. For the structuralists, this was precisely the false problem, and the "depth" of experience could in fact only be an effect of structures which are not themselves experiential. It is in this context that in 1959 Derrida asks the question: must not structure have a genesis, and must not the origin, the point of genesis, be already structured, in order to be the genesis of something?[1]

  1. ^ Jacques Derrida, "'Genesis' and 'Structure' and Phenomenology," in Writing and Difference (London: Routledge, 1978), paper originally delivered in 1959 at Cerisy-la-Salle, and originally published in Gandillac, Goldmann & Piaget (eds.), Genèse et structure (The Hague: Morton, 1964), p. 167:

    All these formulations have been possible thanks to the initial distinction between different irreducible types of genesis and structure: worldly genesis and transcendental genesis, empirical structure, eidetic structure, and transcendental structure. To ask oneself the following historico-semantic question: "What does the notion of genesis in general, on whose basis the Husserlian diffraction could come forth and be understood, mean, and what has it always meant? What does the notion of structure in general, on whose basis Husserl operates and operates distinctions between empirical, eidetic, and transcendental dimensions mean, and what has it always meant throughout its displacements? And what is the historico-semantic relationship between genesis and structure in general?" is not only simply to ask a prior linguistic question. It is to ask the question about the unity of the historical ground on whose basis a transcendental reduction is possible and is motivated by itself. It is to ask the question about the unity of the world from which transcendental freedom releases itself, in order to make the origin of this unity appear.

this is a good example of why a seondary source would be preferable Semitransgenic (talk) 20:21, 27 December 2008 (UTC)

I just pulled up the Dasenbrock article you cited. Of the 75 pages of Derrida's response to Searle that it summarizes, it is summarzing approximately... 10 pages of it. The vast majority of Derrida's response to Searle and engagement with him is not touched by that essay. This is typical - the problem you run into is that simply summarizing existing scholarship (such as Derrida's) is not considered meaningful scholarship for the most part, and won't get published in a journal. So people summarize the parts they're working with in their arguments, but not the rest. Sure, perhaps if we stitched together a dozen or so articles we could cover the bulk of Derrida's engagement with Searle, and thus, by summarizing all dozen piecemeal, create a patchwork summary of Derrida. I would suggest, however, that this is a staggeringly stupid way to approach the problem, and is unlikely to produce a satisfactory result. Phil Sandifer (talk) 22:54, 27 December 2008 (UTC)
If no reliable sources discuss the rest of the debate between Derrida and Searle, then it is probably not notable enough for us to discuss it. Phil, I have the idea that you are trying to write high end scholarly articles, which is not Wikipedia's goal. Simplify. Blueboar (talk) 01:28, 28 December 2008 (UTC)
I think that's part of the problem here. Scholars do original research; that's what makes them scholars. Wikipedia does not. Jayjg (talk) 01:30, 28 December 2008 (UTC)
Oh don't be silly. Plenty of sources discuss the debate between Derrida and Searle. The problem is that they do not offer summaries of the debate. They offer discussion, analysis, critique, response, and expansion, but their primary purpose is not to summarize it. But part of presenting an encyclopedic overview of the debate is summarizing it. Phil Sandifer (talk) 02:02, 28 December 2008 (UTC)
And Wikipedia should summarize the secondary sources' discussions. Jayjg (talk) 02:58, 28 December 2008 (UTC)
I wasn't being silly... I don't know the sources as well as you do, and you made it seem as if there was a dirth of reliable secondary sources. If there are lots of sources, then what is the problem? Just summarize what these sources say and cite them. Blueboar (talk) 02:13, 28 December 2008 (UTC)
Sure, I agree, we should do that. But look, an article on, say, Of Grammatology should probably summarize the arguments of the book. It should summarize all the secondary sources too, but a summary of the book is actually a pretty important part of the article. And I'm less than thrilled with the quality of secondary sources for the purposes of summary. For other stuff, they're great. They're absolutely useful, and we absolutely should use them. But for summary, they're not so good. Phil Sandifer (talk) 02:40, 28 December 2008 (UTC)
Phil, many of the published secondary sources are by individuals who are considered experts, what you may feel about what they have to say regarding Derrida is immaterial, they are qualified to speak, they are published, and this is not the place to take up the defence of one person or another because you might happen to be a fan. In terms of Derrida/Searle, what are the published views? what is the majority view? what is the minority view? in general, what views have the experts offered on the debate? these are the issues that need to be considered, there is no need to go directly to Derrida's writings, in an attempt to offer your view of Derrida's positon, that would be OR. Semitransgenic (talk) 10:47, 28 December 2008 (UTC)
Sure. But all of that still leaves out a key aspect of the topic - summary. In the Derrida/Searle debate, this is particularly bad, because Searle, as a secondary source on Derrida, can be summarized, but Derrida responding directly and explicitly to Searle, cannot. Thus, effectively, Searle gets free reign in the article. Phil Sandifer (talk) 13:21, 28 December 2008 (UTC)
You seem to be going around in circles, there are two cites for sources external to the Derrida/Searle above, there are many more, and you appear to know this, but you find them all unsatisfactory. Maybe you could write a paper that presents your position, you can get it peer reviewed, get it published, then you can cite it. BTW have you looked at WP:POINT? Semitransgenic (talk) 14:35, 28 December 2008 (UTC)
It's not so much that I am going around in circles as that I am continuing to ask that you answer the point I'm giving. Let me try asking the question directly - given that secondary sources critical of a subject can be summarized with no restrictions regarding specialist knowledge, how do we adequately summarize a subject's response to that criticism? Answer this in the general case please, not the specific of Derrida, where it is clear that you do not know the specifics and are not capable of making a useful statement on the matter. Phil Sandifer (talk) 15:16, 28 December 2008 (UTC)

(outdent) WP:NPOV states:

  • Articles should provide background on who believes what and why, and which view is more popular; detailed articles might also contain evaluations of each viewpoint, but must studiously refrain from taking sides.
  • But on such pages, though a view may be described, the article should make appropriate reference to the majority viewpoint wherever relevant, and must not reflect an attempt to rewrite majority-view content strictly from the perspective of the minority view.
  • Keep in mind that in determining proper weight we consider a viewpoint's prevalence in reliable sources, not its prevalence among Wikipedia editors.
  • If you are able to prove something that few or none currently believe, Wikipedia is not the place to première such a proof.

how do we adequately summarize a subject's response to that criticism? I can only speak of the example here, and in this case, it is simply beyond the scope of the article. An article tells us about someones ideas, we have a reception section, you can say there was a debate, you can use secondary sources to describe it, they will do the job adequately, you can then create a content fork to an article called Derrida Searle debate and mull over the details there if you so wish. Semitransgenic (talk) 15:44, 28 December 2008 (UTC)

I'm sorry, is your answer really that the correct problem is not to summarize criticisms of a subject either? Brilliant. Creating a NPOV encyclopedia by providing no information whatsoever. It's a wonder we didn't think of it sooner. Phil Sandifer (talk) 15:46, 28 December 2008 (UTC)
and I'm sorry that the many reliable secondary sources available to you do not reflect your POV. I have no more to offer on this issue. Cheers. Semitransgenic (talk) 15:58, 28 December 2008 (UTC)
No more? You've offered nothing on this issue, continuing to ignore the point I'm making - Searle's criticism is already a reliable secondary source on Derrida. The problem is that by marginalizing Derrida's self-defense, we introduce a NPOV problem - one that is not solved by going to secondary sources on Derrida's self-defense - it still leads to the problem that the subject of an article's response to criticism is a significant viewpoint on its own merits. Phil Sandifer (talk) 16:13, 28 December 2008 (UTC)

Use of letters in biographical articles

I was recently involved in a discussion on the wiki-en-l mailing list (one related to the topics Phil Sandifer has mentioned above) and I raised the topic of letters by a person used in biographical articles on Wikipedia about that person. What is the correct way to use such letters? This policy mentions letters as primary sources, and they pretty obviously are. The examples I had in mind were J. R. R. Tolkien and John Buchan (whose article seems to have ended up at an aristocratic title, naming conventions <mutter, mutter>), but none of the stuff in Buchan's article is sourced or referenced to letters he wrote. The Tolkien article, in contrast, draws heavily on the Letters volume published by Carpenter (the official biographer) as well as the official biography. Actual letters are mentioned at least seven times, and numerous statements in the article are sourced to letters. Letters are, indisputably, an important part of biographical writing, but should it be us reporting others writing about the letters, or can we refer directly to the letters, often (in the case of the prolific letter writers) published posthumously in a volume or several volumes? I will try and dig up some other examples. Churchill, I think was one. Yes, he was. See Winston Churchill, where his wife (I think) published a volume of their letters posthumously. It is given in the "secondary sources" section: "Soames, Mary, ed. Speaking for Themselves: The Personal Letters of Winston and Clementine Churchill (1998)." But this work is cited numerous times. So in what sense can such letters be cited for:

  • (a) Descriptions of events in the lifetime of the person?
  • (b) Opinions expressed by that person?

I have some opinions on this, but wanted to get some responses here first. Carcharoth (talk) 00:52, 29 December 2008 (UTC) disclaimer

As it stands, the letters can be used for either, so long as they are used only for obvious claims. The exact phrasing here is quibbled with - there's a clause about "descriptive" and "interpretive" claims that I think is unclear and requires fixing, but I'm, for the time being, working on the "no specialist knowledge" clause instead of that one. But the general gist is that if the letters clearly say something, it can be said in the article that the letters clearly say it. Phil Sandifer (talk) 01:11, 29 December 2008 (UTC)

Editing warring in place of consensus?

It needs to stop. Now. Professor marginalia (talk) 02:51, 29 December 2008 (UTC)

Erm... this is an interesting thing to post when fully 15 posts of discussion have taken place since the last time anyone reverted the page, and indeed, no reversions have happened since your last post to the talk page. Phil Sandifer (talk) 02:54, 29 December 2008 (UTC)
I'm posting the reminder in the event anyone had it in mind to simply let the necessary hours pass until their next revert to avoid the 3rr restriction. There are other reasons than the threat of a 3rr block to stop this now. Professor marginalia (talk) 03:01, 29 December 2008 (UTC)
Phil, 15 posts of people disagreeing with each other does not equal consensus. In this case, given the strong objections to changing the policy that have been expressed, I would say that there has to be an affirmative, and clear indication that most of the editors who have objected to changing the policy no longer do so. Blueboar (talk) 14:08, 29 December 2008 (UTC)

The problem, wider

Critical Inquiry, one of the top journals in the humanities, publishes frequently a section called "Critical Response," in which a writer pens a response to a previous essay in the journal, and the writer of the original essay responds in turn.

I do not know how many such sections have appeared in the 30+ years of the journal - they were included in almost every issue in the early days of the journal, and still appear from time to time. Some of these are landmark debates about which numerous secondary sources exist - Derrida engaged in a particularly famous one over the legacy of Paul de Man, another raged so widely as to be made into a stand-alone book called Against Theory, and a third, among Wayne Booth, J. Hillis Miller, and M.H. Abrams, is one of the most cited exchanges on deconstruction. Others are less widely cited.

To pick a lone example from the latter category, the Autumn 1995 issue of the journal has a response by Richard Shusterman to an article from the Summer 1994 issue by Tim Brennan, followed by a response from Brennan. Although we do not have an article on Brennan, he is cited widely enough that he meets WP:N. So let us imagine the article Tim Brennan.

Shusterman's response to him, being published in a prestigious journal, is surely a notable perspective, and thus would be included in Tim Brennan. But Brennan's rejoinder to Shusterman is a primary source for Tim Brennan. And as Brennan's work is specialized, there are obvious problems in using his response - simply put, we can't summarize it as thoroughly as we can Shusterman's equally specialized and technical critique of Brennan's work. (Add to this the fact that Brennan's original essay from 1994 actually quotes Shusterman extensively, and so this problem actually exists for both Tim Brennan and Richard Shusterman). But while Brennan's article is cited 19 times or so that Google Scholar can find, one is hard pressed to find a lot of secondary sources on this particular exchange, as, unlike the Derrida/Searle exchange, it is not so significant as to potentially deserve its own article.

This is a NPOV problem. One side of a debate gets disproportionate representation, and a far easier standard of evidence for inclusion. The other side is given a short shrift amounting to "Tim Brennan responded to these critiques." NPOV? No. Not even a little.

There are dozens of near identical cases in this journal alone, which is one journal in one academic field.

See the scope of the problem now? Phil Sandifer (talk) 05:12, 29 December 2008 (UTC)

Taking the Brennan/Shusterman example, could you give us some flavor of the "novel interpretation" required in citing Shusterman's response? Professor marginalia (talk) 05:44, 29 December 2008 (UTC) Or is it Brennans? Professor marginalia (talk) 05:45, 29 December 2008 (UTC)
In neither case is novel interpretation required - realistically, anyone who understands a given paragraph of either response at all is going to come to the same conclusion about what it means. The writing is no more ambiguous than the average New York Times article - it's just hard, such that not every reasonable, educated person who reads it would understand it.
The problem is that, as policy stands, that would mean that even the very obvious conclusions that anyone who understands the work at all would come to would be forbidden from inclusion. Phil Sandifer (talk) 13:25, 29 December 2008 (UTC)
If there are "very obvious conclusions that anyone who understands the work at all would come to"... then 1) it should be easy to summarize them in a way that the average reader of Wikipedia will be able to verify, and 2) it will be very likely that a secondary source will have discussed these obvious conclusions... and thus we can cite secondary sources. If neither of these are true, then the conclusions are not that obvious in the first place, and there is a very good chance that they are novel and OR. Blueboar (talk) 14:17, 29 December 2008 (UTC)
I haven't been following this discussion but I would like to comment on item 2 in the remarks that Blueboar just made. Something that is obvious for people familiar with the subject, probably won't be mentioned in secondary sources. But for some readers of the Wikipedia it may not be obvious and should be mentioned by editors to clarify the article.
I've read the item 2 argument before by others and I believe it's fallacious because obvious is relative. Something could be obvious to authors of a secondary source and the specialist audience intended to be reached by the author, and not worth including in that secondary source. On the other hand, it could also be obvious and non-controversial to the editors of a Wikipedia article, yet not obvious to most readers of the article and thus beneficial to the article. --Bob K31416 (talk) 14:56, 29 December 2008 (UTC)
Yes. Something I've pointed out from time to time, and that we need to be aware of (not necessarily change our policy around, but be generally aware of) is that publication is selective process that is not selective on merit alone - it's also selective on commercial grounds. Simply put, no academic press will publish something with no chance of selling. Similarly, journals do not simply publish the best material, but the material that seems of most interest to their readers, creates a generally interesting table of contents, etc. Publishing is a commercial process that does not have a 1:1 correspondence with merit or importance. And we need to, in constructing a guide to all knowledge, recognize that there will always be gaps in the secondary source record that need to be filled. The challenge is figuring out how to fix them without OR. Phil Sandifer (talk) 21:12, 29 December 2008 (UTC)
This all comes down to what I call the "Oh yeah? Says who?" question. For any statement we make, we need to both ask and be able to answer that question. If the answer is... "A Wikipedia editor says so" then the statement is OR. If the answer is "This reliable source says so (citation)" then it is not OR. It does not matter how "obvious" the statement is... if it can not be cited, it is OR. (note: there is a difference between can not be cited, and is not currently cited.) Blueboar (talk) 15:38, 29 December 2008 (UTC)
The objection is that there is a NPOV conflict. The policy says that "interpretations" of a primary source are limited to secondary sources, and absent secondarily sourced "interpretations", primary sources can be used but are limited to descriptive claims only--claims which are easily verified by nonspecialists. How this is a NPOV problem remains unclear to me in this example also. It's not even clear why any one of two individuals embroiled in fight with each other would be a "primary source" and the other a "secondary". Professor marginalia (talk) 16:34, 29 December 2008 (UTC)
Blueboar, what you are saying doesn't match current policy. Current policy says that we may summarize criticism, but we may not summarize the original author's reply to the criticism. If what you are saying were true, then either a summary is OR, and we can't summarize either one, or a summary is not OR, and we could summarize both of them. Neither of these is what we do now. 208.89.102.112 (talk) 16:44, 29 December 2008 (UTC)
No, it says we use "interpretation" of a primary source when it's cited to a secondary source, but a primary source alone can only be a cite if it is straightforwardly described--and this includes a straightforward summary. Professor marginalia (talk) 16:53, 29 December 2008 (UTC)
However, the line also says that even straightforward description is not OK if that description requires specialist knowledge. What this comes down to is that there are actually specialist topics in the world. I disagree strongly with Blueboar's suggestion that these can somehow be simplified - there are actually complex topics, and actually complex sources, and not all general audiences are going to be able to just seamlessly sort that out for us. Phil Sandifer (talk) 21:12, 29 December 2008 (UTC)

I've tried to satisfy myself what the problem is using this example of Shusterman and Brennan, but it only highlights the importance of policy to me. The overwhelming majority of the Richard Shusterman article is either unsourced, synth, or consists of unsourced interpretations drawn from from primary sources--ie original research. Professor marginalia (talk) 17:04, 29 December 2008 (UTC)

For what it's worth, I agree, that is a poor article. I was unaware we had a Shusterman article when I picked the example, actually. But again - dozens more examples. We can go with the 1988 debate between Frank Lentricchia and three other scholars instead. Or any number of other debates in Critical Inquiry alone. Phil Sandifer (talk) 20:56, 29 December 2008 (UTC)

Specialist knowledge clause violates NPOV

The more I look at this, the more I am convinced that the prohibition on summarizing primary sources that require specialist knowledge is a violation of WP:NPOV. Here's an example of the problem:

here is a very famous debate that went on over three essays between Jacques Derrida and John Searle. Searle's attack on Derrida - the second of these three essays - is clearly a significant point of view on Derrida, and NPOV requires reporting it. Furthermore, as it is a secondary source, it can be summarized with impunity, as there is no specialist knowledge rule on secondary sources.

Derrida's response is equally self-evidently a significant view, as Derrida responding to significant attacks is clearly something we need to report under NPOV. But Derrida's essay - which is long and technical - requires specialist knowledge to explain. Thus it cannot be summarized directly.

The real problem is that this is not an isolated case - any time there is a debate in an advanced, specialist topic, we are going to run into this problem - criticism can be summarized directly, while the subject's response to the criticism cannot. This is a flagrant violation of NPOV, stacking the deck in every single specialist topic on Wikipedia. This is positively disastrous for BLPs - poison pen critics can sandbag a subject left, right, and center, and if the matter is technical, the subject's responses often cannot be included in articles, or, if they can, can only be included inasmuch as they are discussed by others, while the critics are under no such restriction. Absurd, and clearly a violation of NPOV.

The specialist knowledge clause is in violation of WP:NPOV - the one of our content policies that is mandated as a Foundation issue. NPOV is by far the most important one of our content policies - the Foundation has shut down projects for failing to follow NPOV. NOR, while a fundamental policy, does not supersede NPOV, nor can it.

This clause absolutely has to be removed. Phil Sandifer (talk) 02:02, 28 December 2008 (UTC)

Um... Phil... the policy does not have a "prohibition on summarizing primary sources that require specialist knowledge"... look at the wording again... (I will put the key passage in bold)
  • Our policy: Primary sources that have been published by a reliable source (for example, by a university press or mainstream newspaper) may be used in Wikipedia, but only with care, because it is easy to misuse them. Any interpretation of primary source material requires a reliable secondary source for that interpretation. Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is easily verifiable by any reasonable, educated person without specialist knowledge. For example, an article about a novel may cite passages from the novel to describe the plot, but any interpretation of those passages needs a secondary source.
The phrase "specialist knowledge" refers to the person who might want to verify the descriptive claims we have written. It does not refer to the primary source. In other words, anything we say about a primary source must be verifiable by a person who does not have a specialist's knowledge. Essentially that sentence defines who we think Wikipedia's audience is: Reasonable, educated people without specialist knowlege. Blueboar (talk) 02:31, 28 December 2008 (UTC)
So if the claim would not be understood by someone without specialist knowledge, it cannot be made. Which means that, for instance, Derrida's response to Searle cannot be summarized (as there is little that can be gleaned without specialist knowledge), even as Searle's attacks on Derrida can be summarized. As would be true for any dispute on a technical matter - secondary sources requiring specialist knowledge can be summarized freely. Primary sources responding to them cannot be. This violates NPOV on every single dispute over a technical matter. Phil Sandifer (talk) 02:37, 28 December 2008 (UTC)
No, that is not what the passage says... it says that the reader (that reasonble, educated person without specialists knowledge) has to be able to easily verify the accuracy of what we write (the descriptive claim) about the source. It says nothing about whether the reader understands the source.
The passage is all about what WE write... not what is in the source. Blueboar (talk) 02:59, 28 December 2008 (UTC)
If the reader does not understand the source, they seem to me unlikely to successfully verify the accuracy of a summary. Phil Sandifer (talk) 03:01, 28 December 2008 (UTC)
Then you didn't do a good job at summarizing it. The fault is with the editor, not the source. Blueboar (talk) 03:04, 28 December 2008 (UTC)
If I don't understand the source, I can't verify the accuracy of a summary. Even a very good summary. Phil Sandifer (talk) 03:09, 28 December 2008 (UTC)
I seriously doubt this would ever be a problem, a good summary would help to explain the primary source (while at the same time would avoid violating this policy). But... OK, let's assume that there is some hypothetical source that is so complex that the average reader would never be able to verify the accuracy of a descriptive claim, no matter how well written. In this exremely rare case, the solution would be for the article writer to turn to secondary sources completely, and not use the primary source at all. Blueboar (talk) 03:28, 28 December 2008 (UTC)
If the work is technical and requires specialist knowledge, one is hard pressed to verify a summary of it. I mean, how would one go about this? How would one go about verifying the accuracy of a summary when they do not understand the work it is summarizing? A good summary, I suppose, will appear to be reflected in the work - but no shortage of bad summaries will as well. If the reader does not understand the source prior to a summary, I am disinclined to trust them to judge good summary from bad. How do you propose that this process would work? Phil Sandifer (talk) 03:43, 28 December 2008 (UTC)
I have no idea... I have never come across a primary source that I could not write a basic descriptive summary of. But if I did, I problaby would not even try to write a descriptive summary of the primary source. Instead, I would focus on what the secondary and tertiary sources have to say. Blueboar (talk) 03:53, 28 December 2008 (UTC)
Well, this is unsurprising, as I doubt you regularly edit in areas you don't know about. I assume, however, that you have come across primary sources that others could not write a basic descriptive summary of. And all the same, while that might be what you would do, this does not erase the fact that permitting free summary of Searle attacking Derrida as he is a secondary source while restricting summary of Derrida's response to Searle is a grotesque violation of NPOV.
If nobody can explain how on earth it does not undermine NPOV to treat secondary sources critiquing a subject under a looser standard than the subject's defense of his or herself, and particularly how it does not cause massive BLP problems, then I do not see any option beyond deleting the words "without specialist knowledge" from the policy. Phil Sandifer (talk) 04:00, 28 December 2008 (UTC)
Phil, as stated already, there's a multitude of usable secondary sources, represent those. Do not present a defense based on your interpretation of Derrida's repsonse, that would be OR. Also, I've seen a number of introductions to both Derrida and Deconstruction that are very accessable to the non technical reader, so I really do not get your point here. As you have been told already, a wikipedia article is not an acadmeic paper. Please do not adjust policy without concensus. Semitransgenic (talk) 11:02, 28 December 2008 (UTC)
It is a violation of NPOV to treat the two sides of an argument differently. This is not about Derrida alone. It violates NPOV in a fundamental fashion if one significant side of an argument can be summarized freely, and the other can't. Criticism, always being a secondary source, gets in under laxer rules than self-defense. That's inappropriate. Phil Sandifer (talk) 13:18, 28 December 2008 (UTC)
Actually, no, there is no NPOV violation, since the points of view for both sides can be discussed using secondary sources. Blueboar (talk) 15:25, 28 December 2008 (UTC)
Are you even reading what I'm saying? OK. Let's try this again, in the general case.
Person A is a scholar working in a specialized area. Person B writes an essay criticizing Person A's work, and publishes it in a peer-reviewed journal. Person A responds, also in a peer-reviewed journal. Both of these works are highly technical. For the article Person A, Person B's essay is a significant criticism that should be covered, and, as it is a secondary source to the subject of Person A, can be summarized directly. Person A's response, however, is a primary source on the subject of Person A, and so is subject to more stringent requirements. Thus criticism of Person A is included under easier standards than Person A's defense. This is a NPOV problem. Phil Sandifer (talk) 15:31, 28 December 2008 (UTC)
relative to the other secondary sources that have commented on the exchange between A & B, Person A might suddenly be the sole representative of a minority view, in which case it would not require coverage. Semitransgenic (talk) 15:54, 28 December 2008 (UTC)
I'm sorry, your argument isn't really that the subject of an article's views are not significant to the article, is it? Phil Sandifer (talk) 15:55, 28 December 2008 (UTC)
I'm with User:Blueboar on his reversal of your edit Phil Sandifer. I think that one need to read the passage that you edited in the context of the content policies in general and not in isolation, as doing that tends to throw up the sorts of concerns that you are expressing. --PBS (talk) 16:03, 28 December 2008 (UTC)
I am actually thinking about it in general - which is part of why I find this clause so baffling. It is clearly a clause that applies primarily to articles on specialist subjects, which are exactly the articles where it introduces NPOV problems on every single article it is applied to. I cannot think of any other purpose it serves that is not covered by another policy. So what, exactly, good does it do? Besides winning a content dispute on animal rights over three years ago, which, I hasten to reiterate, was all it was originally designed to do. Phil Sandifer (talk) 17:55, 28 December 2008 (UTC)
Why not avoid the entire issue? Use a reliable secondary source to discuss the entire debate... and avoid using either the critique or the response directly. Blueboar (talk) 16:07, 28 December 2008 (UTC)
Again, your solution here appears to be to avoid the reliable secondary source of Person B's article. That's also a NPOV violation. Phil Sandifer (talk) 16:11, 28 December 2008 (UTC)
No, because both person A's views and person B's views can be discussed using person C's discussion of their debate. Neutrality is preserved. There is no rule that says a specific source has to be used to present a given view. If that viewpoint can be presented using a different source, then NPOV is complied with.
Finally, in those extremely rare situations where you must use a pimary source directly, and the only way to do this would end up violating NOR... go ahead and do so. You can always invoke WP:IAR if you really think it is in the best interests of the project. Blueboar (talk) 16:19, 28 December 2008 (UTC)
This seems like a very silly way to deal with the problem - and we'd be hard pressed to find a good excuse to remove a section on Person B's work that is sourced to Person B. That makes me think that this is not an extremely rare situation - in fact, I would suspect that this is going to be the case any time that there is any criticism of someone who works on a specialist or technical matter. Phil Sandifer (talk) 16:47, 28 December 2008 (UTC)
So don't remove B's work... just use C to discuss A's view point... the point is, there are options available. All viewpoints can be represented neutrally, so there is no NPOV violation.
In any case, I'm done here... you have recieved a clear answer from multiple editors that this section of the policy has consensus. At this point further discussion would indeed be WP:POINT. We have presented several options that can be taken when the situations you describe occur. The fact that you don't like any of these options is your problem, not that of the policy. Blueboar (talk) 17:19, 28 December 2008 (UTC)
Please don't invoke WP:POINT to mean "something I don't agree with" - it doesn't make your argument useful or convincing, quite the opposite - David Gerard (talk) 18:45, 28 December 2008 (UTC)
It is not NPOV to allow one side of the debate unfettered access to summary while attempting to reconstruct the other side from secondary sources that may or may not exist! Since your responses have amounted to "la la la can't hear you," or bizarre assertions that this is not a NPOV problem, I can only conclude that you are simply not putting effort into this discussion, or do not care. In any case, I am removing the "no specialist knowledge" clause as an NPOV violation, as this policy cannot trump NPOV. If you can come up with another plausible fix, feel free to propose it, but it is simply not acceptable to mandate for a wide class of articles that criticism can get in on an easier standard than the subject's defense. Phil Sandifer (talk) 17:25, 28 December 2008 (UTC)
If you want a neutral point of view about deconstruction, sample the secondary sources. This summary of Deconstruction for Beginners, believe it or not, quite arguably represents the broad consensus (roughly, the NPOV) about Derrida's work, not just among the peanut gallery or armchair philosophers but also among numerous skilled philosophers and analysts. The publisher's pitch for this widely sold, insightful, "plain-English" commentary on deconstruction reads:

Deconstruction is so labyrinthine that it has become the monster that murdered philosophy. When Jacques Derrida, the father of deconstruction, uses buzz-words such as "phallogocentrism" and "transcendental signified," humanities students and aspiring philosophers may get weak in the knees."

A more sympathetic view is offered in Deconstruction: Theory and Practice, by Christopher Norris. Norris' perspective is consistent with the historical perspective of the philosophy leading up to Derrida that's presented in Deconstruction in Context: Literature and Philosophy, Mark C. Taylor (Editor). Both of these approaches, and others too, present Derrida's deconstruction as a kind of whimsical stepchild of French structuralism. (Mark Taylor's book is actually an anthology starting with Kant, I believe, and going through Derrida, but the issue of deconstruction being a response to structuralism, with existentialism in between, is central to Taylor's choice of material and order of presentation.) In other words, the consensus is, more or less, that to take deconstruction seriously and strictly from an analytical standpoint is to completely misunderstand it. Rather, Derrida's deconstruction is philosophically inventive and playful, a reaction to structuralism. That, in extremely rough terms, is the NPOV about deconstruction. Thus, it simply isn't true that because the secondary sources tend, on the whole, to be variously critical of Derrida's work, that somehow this particular WP editorial policy is a problem.
..... Rather, far from violating or diverging from NPOV, this aspect of WP:NOR facilitates WP:NPOV, even in the case of Derrida's deconstruction. ... Kenosis (talk) 18:28, 28 December 2008 (UTC)
For Deconstruction, those are all decent sources. They're even good sources for Jacques Derrida. But they are not sufficient. For Jacques Derrida, no secondary source provides an adequate replacement for the POV of Derrida himself. To cut Derrida's own self-defense out of the article violates NPOV. And secondary sources on Derrida's own self-defense still do not actually introduce that crucial perspective. And again, the issue is a general one - Derrida is one example of a problem that is present for every single article on specialist topics we have. Phil Sandifer (talk) 18:39, 28 December 2008 (UTC)
Phil, please, maybe sleep on this for a few days, perhaps reread the tack of your arguments thus far when you get a chance, re-analyze them in the context of all three core content policies, and come back to it fresh later on. Maybe there's something everybody missed here -- offhand, I personally don't see it. ... Kenosis (talk) 18:43, 28 December 2008 (UTC)
Quite simply: NPOV is the core content policy. NOR and V are ways to get there usefully and reliably. The sentence fragment in question was put in by one editor a few years ago to win a particular content battle - in, I'm sure, the best of good faith, but clearly without considering the consequences. It's clearly and obviously insane when dealing with real world content, and Phil's presenting the reasons why it's so, with examples. If a sentence in NOR violates NPOV, it has to go and there's no two ways about that. Process isn't an end in itself, and in this case it hampers product - David Gerard (talk) 18:48, 28 December 2008 (UTC)
Wrong. Quite simply, the language of the policy on appropriate scope of the use of primary sources has been repeatedly and thoroughly discussed, and it has very broad consensus among experienced WP users who're intimately familiar with the core content policies. ... Kenosis (talk) 19:09, 28 December 2008 (UTC)
yes, fair enough but phil is also trying to alter policy to suit his own adgenda, please consider this before reaching a conclusion. Phil is insistent on ignoring inumerable secondary sources in favour of his reading of a primary source, because he believes the secondary sources are inadequete. Semitransgenic (talk) 18:53, 28 December 2008 (UTC)
No-no-no-no. If "no secondary source is an adequate replacement" for one of Derrida's rebuttal, then the same could be said of "no secondary source is an adequate replacement for primary sources, period." Not all content taken from a primary source is "original research", but certainly "specialist only" analysis, argument, or interpretation of a well-known primary source that nobody else has or can describe "adequately" except for some editor here on wikipedia is exactly the kind of "original research" we should be most careful to exclude. This need not run us into problems on NPOV. The balance and content needed for NPOV is achieved by representing fairly the most significant published sources writing about the subject. If the Derrida/Searle dispute is covered in secondary sources (and it seems to be, by "specialists") then NPOV is achieved by describing what those sources say. Professor marginalia (talk) 19:31, 28 December 2008 (UTC)
Well, from an NPOV standpoint, no secondary source is an adequate replacement for the primary source when we're trying to describe Derrida's POV on Derrida - there's a distinct value to Derrida's own words on that issue that it is not entirely possible to overcome. What is notable is that, for more or less this exact reason, we allow primary sources - there are things secondary sources simply are not adequate for. It's this specialist knowledge restriction that poses a problem - because it effectively re-forbids primary sources in some areas without solving the underlying problems that necessitate primary sources' inclusion on Wikipedia. Phil Sandifer (talk) 17:39, 30 December 2008 (UTC)

A note to Semitransgenic - you may not be the best person to discuss the minutiae of Wikipedia policy if you're unaware of WP:3RR - David Gerard (talk) 19:33, 28 December 2008 (UTC)

Where is there evidence of any consensus to remove the specialist knowledge clause? Professor marginalia (talk) 19:46, 28 December 2008 (UTC)
Where is there evidence of any consensus to violate NPOV in the wording of a lower policy? There's no such thing possible - David Gerard (talk) 20:01, 28 December 2008 (UTC)
The protest that the clause "violates NPOV" is a very weak one thus far, especially with this example cited. And NPOV is not a policy paramount to all the others or we'd welcome Derrida (if he were alive) and Searle to add their own unpublished arguments to these pages. Jimmy Wales would write copy for his own article as well. No such priority exists for putting NPOV over sources. Professor marginalia (talk) 20:26, 28 December 2008 (UTC)
WP:NOR is hardly a "Lower" policy... it is one of the core policies, on an equal footing with NPOV and V. All three have to work in harmony with the other two (and in harmony with other polices and guidelines). The contention here is that you have to violate WP:NOR in order to comply with WP:NPOV and vise versa. I don't think that is the case. There are ways to deal with the situation being discussed here that comply with both policies. There is thus, no need to change this policy, as it does not actually conflict with NPOV. Blueboar (talk) 22:19, 28 December 2008 (UTC)
Please review m:Foundation principles. NPOV is uncontrovertibly a content policy in a way that NOR is not. The Foundation has actually closed projects for failing to adhere to NPOV. NOR really, actually does not have the authority to undermine NPOV. Phil Sandifer (talk) 22:37, 28 December 2008 (UTC)
First of all I must say how nice it was to come across David Gerard's post which I frankly found to be an utter lesson in rudeness and low blow to a legitimate point raised by Semitransgenic and now this whole convoluted nonsense about how NPOV somehow is the highest content policy on Wikipedia and NOR and V help you get there. Uh no, David, and frankly I don't care what civility policies you claim ownership over and feel you can direct others when it's okay to reference them (WP:POINT). For clarification there is a WikiMEDIA foundation issue called NPOV and there is WikiPEDIA's NPOV content policy. The WikiMEDIA article simply states NPOV content policies exist across all the wikis of -news, -books and -pedia. It does NOT state NPOV overrides or somehow relegates all other content policies within a wiki to a lower status. This is what David Gerard tried to propose and I found it embarrassingly incorrect. V, NOR and NPOV are equal content standards on Wikipedia and none of them should override the others, end of story. Tmore3 (talk) 01:30, 29 December 2008 (UTC)
Dancing over the irony of such rudeness being used to complain of incivility with a swift rhetorical trope, I would point out that m:Foundation issues says, and I quote, that "neutral point of view as the guiding editorial principle" is "beyond debate." Not "a guiding editorial principle," but "the guiding." Thus it has a core status that no other policy touches. Neutral point of view, in this case, links to m:Neutral point of view, which should thus be taken as an absolute bedrock principle that all other policies must adhere to. This bedrock principle includes the idea that we must try "to present a fair, neutral description of the facts — among which are the facts that various interpretations and points of view exist." To heavily marginalize a subject's own words in response to criticism of them is clearly a violation of this. Thus this policy is currently in violation of m:Foundation issues. Phil Sandifer (talk) 01:38, 29 December 2008 (UTC)
Ah Phil, sorry if I haven't chosen to spend any time debating in circles with you again but if it's anything like the last discussion where you made the claim that WP:SYNTH should be changed to allow unpublished interpretation of a primary source because "people don't write about obvious things" something tells me I haven't missed all that much. Tmore3 (talk) 03:44, 29 December 2008 (UTC)
If that is how you opt to summarize my previous statements on this policy, it's clear that you've missed just about everything, so why break the streak. Though one wonders, if you're not interested in contributing to the discussion, you're contributing to the discussion. Phil Sandifer (talk) 03:57, 29 December 2008 (UTC)

User:Phil Sandifer you wrote above "It is clearly a clause that applies primarily to articles on specialist subjects, which are exactly the articles where it introduces NPOV problems on every single article it is applied to." This is a problem with assuming that one knows the content of the millions of articles on Wikipedia from the small sample that one has read. Let me give you an example where this rule is very useful: The section "International humanitarian law" in the Wikipedia article "military use of children" discusses Article 77.2 of the Additional Protocol I to the Geneva Conventions. How is a "NPOV problem" introduced by using the ICRC commentary to explain 77.1, because without that commentary how would one know the precise meaning of "take all feasible measures"? (One of the problems with internet blogs is that they do this and make all sorts of claims about international law because they think they know what a treaty says without bothering to read the expert commentaries on the treaty) -- a mistake it is easy to make and hence why the current wording is useful for the interpretation of the treaties that make up international law. For another example see the Genocide article and the meanings of the linked text in the following clause from the Genocide Convention: "... any of the following acts committed with intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such ..." --PBS (talk) 14:00, 30 December 2008 (UTC)

Fair point - thanks for this observation. It helps clarify some of the trickier issues here. Phil Sandifer (talk) 17:39, 30 December 2008 (UTC)

Comparison articles

In an article comparing two entities (football and cricket, for instance, or in this specific case, two empires) is it OR/Synthesis to have sections on aspects of the entitities which just outline those aspects with sources drawn from entity-specific sources? (I think that's pointless, but that is another issue. This came up at Talk:Comparison between Roman and Han Empires in discussing some obvious synthesis in the article, and the editor is now saying "My plan for revamping the article is to use a direct comparison when i can find a good source, e.g., Prof.shiedel's book. But when such a source is not available, I will list information about the two empires without comparing them, as is in several sections currently." My first thoughts are that this is OR, because it is inevitably going to be the editor's selection of information, drawn from sources who may have, for instance, different defintions of a particular concept because they had never meant their particular article or book to be used as a comparison. Hopefully historians who are writing about a comparison of the two empires will always be comparing like for like. For instance, two historians, one writing about the Han Empire, the other about the Roman Empire, might use the phrase 'social mobility' but have different definitions in mind. Hopefully a historian directly comparing the two would use the same definition of social mobility. I'm unhappy about sections that are not drawn from sources directly comparing the two articles, but I can probably be convinced that I'm wrong. This doesn't seem to be clearly covered on the policy page. dougweller (talk) 12:45, 29 December 2008 (UTC)

Hmmm... tricky... Perhaps proximity of discussion is something that needs to be considered. If you structure the article so that you have a section on "social mobility" (with a statement directly contrasting social mobility in each empire, but citing the two sources to support the information as relates to seperate empires) I could see saying that you are setting up a novel juxtaposition, which would be OR. On the other hand, if you structure the article into a secton on the Roman Empire (which talks about social mobility), and a seperate section on the Han Empire (which also talks about social mobility), then I think you are on better footing. While you are still setting up a juxtaposition, it isn't a direct juxtaposition, which might give us more leeway.
It may also help to consider how we would deal with this in some other, related article. What if we flipped the topic... what if we are talking about an article on "social mobility"? Would we consider it OR to include a history section that discuss how social mobility worked in different empires (Persian, Roman, Han, Russian, British, etc.), even if no single source ever directly compared social mobility in these empires?
Finally, I think a lot of this depends on what the sources say. If it is obvious that the two sources are using the phrase "social mobility" to mean different things, then it clearly would be wrong to use them in comparison. If it is clear that they are using the prhase in the same way, then it probably isn't. Blueboar (talk) 14:59, 29 December 2008 (UTC)
The source guideline cautions, "Editors should avoid original research especially with regard to making blanket statements based on novel syntheses of disparate material." One can effectively mislead with their own OR synthesis via these independently derived side-by-sides bookended with a claim or two sourced straightforwardly (eg Shiedel). I've been involved in multiple disputes at WP with conspiracy types who attempt to add "evidence" for their POV this way. Not okay-sources cited must address the topic of the article. If the topic is a "comparison" the sources used for claims must address the "comparison" with limited exceptions. Professor marginalia (talk) 15:15, 29 December 2008 (UTC)
Prof... good point, but that begs the question... what would you say are the limited exceptions? Blueboar (talk) 15:40, 29 December 2008 (UTC)
If sources are used which compare the military strategies used by the two sides in the Civil War, there would be no need to demand in an article about that comparison that even minor, uncontroversial details (what years the Civil War was fought, what state the commanding generals came from, etc) come from a source comparing the two's military strategies. These are the sorts of exceptions that I have in mind. Professor marginalia (talk) 15:55, 29 December 2008 (UTC)
Most of the sections in Comparison between Roman and Han Empires are just 'here's stuff about the Roman military' 'here's stuff about the Han military' with (at least in one section but not most) something in the last paragraph from a source which compares the 2 empires. dougweller (talk) 17:42, 29 December 2008 (UTC)
I think we agreed on the OR noticeboard that that particular article was full of OR and needed a serious re-write. So I don't think it makes a good example. I take this discussion to be more along the lines of... "so where do we draw the line?" If we are writing the hypothetical article "Comparison of X and Y", and we have sources that discuss an aspect of X and sources that discuss a similar aspect of Y... to what degree can we set up a comparison ourselves? Given WP:SYNT, I think we need to draw the line up front. We need a source that compares this aspect of X and Y directly. Blueboar (talk) 17:55, 29 December 2008 (UTC)
Exactly. Ironically, a sentence in the lead paragraph of Comparison between Roman and Han Empires promises, "Each empire brought the provinces of their domains together into regimes of unprecedented scale and thereby enhanced the integration of local worlds into a common legal and cultural framework. Because of this, they have often been compared." With that the thesis, in this or any comparison-type article, then each key cite must speak to the comparisons already made by which authorities. It's not a welcome mat for DIY comparisons. Professor marginalia (talk) 18:03, 29 December 2008 (UTC)
Yeah, that was the first problem sentence I noted in my review of that article, and I stuck a big fat {{who?}} tag on it. But, let's try to get away from that article. It seems that Doug, Prof and I all agree that in order to compare X and Y, you do need cite source that directly compare X and Y, and that do so in the context of each sub-topic you use in your comparison. Does anyone disagree, or is that settled? Blueboar (talk) 18:49, 29 December 2008 (UTC)
I believe that to compare the two articles, you should view WP:SYNT. It says "The best practice is to write Wikipedia articles by taking information from different reliable sources about a subject and putting those claims on an article page in our own words, yet true to the original intent — with each claim attributable to a source that explicitly makes that claim."

This policy sentence, i believe, states that you only have to have sources that support the INDIVIDUAL claim. Regarding professor's solution, i believe that is a viable solution. However, I do not know how to make sub-sections, so I will need assistance on that one. Also, regarding the definition of terms used, I believe the definition of terms used is teh same in each case.Teeninvestor (talk) 17:45, 30 December 2008 (UTC)

It does appear we need to think about making the situation with comparisons explicit so that we don't get people thinking that it is ok to do what Teeninvestor thinks is ok to do. dougweller (talk) 20:19, 30 December 2008 (UTC)

Small change

Just a quick note, I changed, in the "using sources" section, the note that passages that are open to interpretation should be avoided to "passages that are open to multiple interpretations." I do not think this changes the meaning of the line - it merely wards against the description/interpretation problem I've been discussing (and will eventually get back to more substantively when we figure out this specialist knowledge issue). But if anyone thinks this changes the meaning, please feel free to revert - I won't object or fight over it. It just seemed a quick and probably uncontroversial fix to deal with the larger issue there. Phil Sandifer (talk) 17:33, 30 December 2008 (UTC)

New footnote at end of first paragraph

In a recent edit that was deleted, I tried to add the following footnote to the end of the first paragraph of WP:NOR:

"This rule does not, in general, refer to statements that are non-controversial and easily reducible to elementary deductive logic. See also, WP:Verifiability#Burden_of_evidence"

I was inspired by footnote 2 at the end of the second paragraph of WP:SYN, which I used as a model and precedent:

"The rule against 'A and B therefore C' does not, in general, refer to statements A,B and C that are non-controversial and easily reducible to elementary deductive logic. See also, WP:Verifiability#Burden_of_evidence"

It seemed that the same type of reasoning that led to the acceptance of footnote 2 applied to the new footnote. The important requirement of "non-controversial" and the application of WP:Verifiability#Burden_of_evidence should prevent misuse of this footnote, as it prevents misuse of footnote 2. I would be interested in comments on this matter. Thank you. --Bob K31416 (talk) 15:37, 20 December 2008 (UTC)

Hey Bob, replied to your earlier point regarding this but am fine bringing it up here. For reference I removed the original footnote that was included in September in hopes of finding some clear consensus for it. I could not find that this eventually developed so for now have removed it for now to avoid any confusion. Tmore3 (talk) 15:46, 20 December 2008 (UTC)
Within the context of WP:SYNT, I think the foot note is apt... if source A states that all canines have tails, and source B states that all dogs are canines, we can use basic logical deduction to synthesize the conclusion that all dogs have tails. However, I am much less certain that allowing basic deductive logic is apt in other potentially OR situations. Blueboar (talk) 15:55, 20 December 2008 (UTC)
Hi Tmore3, It's good to hear from you here and on your talk page. The previous discussion of footnote 2 may be hard to find and I appreciate your efforts in looking for it. I too had some trouble finding it. You can find discussion of footnote 2 here. Apparently the footnote survived the rigors of that discussion so it seems that it has been accepted. Regards, --Bob K31416 (talk) 16:00, 20 December 2008 (UTC)
I read through that as well as the discussions that occurred previous and latter and was not able to find any consensus for keeping it other than one editor wanting to keep it in at the end of that discussion to see if any further work could be done on it. My main problem is that it it opens up the opportunity to use this as rationale to conduct all sorts of conclusion drawing. If source A makes one statement and source B makes another statement, and a wikipedia editor uses those two sources to make statement C, I don't see how this is anything but original research. I admit some may view this approach as "hardline" but the ramifications to me as we just saw with that footnote being used as precedent to allow additional caveats to OR is in my opinion not worth the trouble it would cause. Tmore3 (talk) 16:10, 20 December 2008 (UTC)
I understand your concern but we should also be concerned with deleting footnote 2 that has been accepted as part of WP:NOR for several months and after lengthy discussion. Just as I accepted your reversion of my edit and I understand the need for discussion, shouldn't the same apply to reversion of your recent edit? There should be a relative amount of stability for something as important as WP:NOR .
Let's wait until there is a reasonable amount of discussion before removing an important part of WP:NOR that has survived for several months.
Re "My main problem is that it opens up the opportunity to use this as rationale to conduct all sorts of conclusion drawing." That is not an argument regarding whether or not it is correct. Please note the safeguards against drawing all sorts of conclusions: non-controversial and WP:Verifiability#Burden_of_evidence.
Re "If source A makes one statement and source B makes another statement, and a wikipedia editor uses those two sources to make statement C, I don't see how this is anything but original research." Editors are always interpreting references, bringing together information from sources, and summarizing articles and groups of articles. This is original reasearch but it is OR that is acceptable. None of these interpretations, bringing together information, or summarizing can be found explicitly in some secondary source yet it is accepted practice for Wikipedia. Often contributions are controversial, as evidenced by discussion on many talk pages. The situation described in footnote 2 should be even more acceptable than those controversial edits that routinely occur at articles, since a precondition in footnote 2 is that the situation be "non-controversial". This is a tough requirement. Can you imagine how difficult it would be to make contributions to the Wikipedia if that same requirement of "non-controversial" was applied to all contributions?
Re "I admit some may view this approach as "hardline" but the ramifications to me as we just saw with that footnote being used as precedent to allow additional caveats to OR is in my opinion not worth the trouble it would cause." - Please see my response to the first item that I quoted.
I appreciate your ability to discuss these matters and I think that it will serve you well in the discussion of this issue, but as I mentioned before, the proper approach is discussion when an edit such as yours is controversial, instead of summarily deleting a 3-month old part of WP:NOR. Thus, I feel that the proper course is to put footnote 2 back into WP:NOR for now and I have done that. I hope that after reading my above discussion you are not offended and please note that I have not restored my new footnote, and only restored footnote 2 that existed previously. Regards, --Bob K31416 (talk) 17:42, 20 December 2008 (UTC)
Just to make sure that my view on this does not get lost in the conversation between Bob and Tmore, let me reitterate... I agree with leaving the footnote in the WP:SYNT section... but I disagree with adding a similar statement to the first paragraph. Much more discussion is needed before we do that. Blueboar (talk) 18:49, 20 December 2008 (UTC)

Bob, I would take issue with your assertion that consensus was reached simply because a footnote with significant voiced opposition was inserted 3 months ago in a 5 year old policy page with the caveat of seeing if it could be improved. I will not revert for now for the sake of this discussion but the language defintely needs some improvement.

Your statement "removing an important part of WP:NOR" actually illustrates my point precisely. That footnote never was part of the NOR policy body itself and as noted in the discussion you referenced earlier, the editors who supported its inclusion as a footnote were vehemently opposed to it as part of the policy itself when discovered it had creeped into the body of WP:SYN. Again there is no problem summarizing and synthesizing a variety of sources in one article in fact it's incumbent upon the editor to do this to construct a good article. However when you start using sources to draw your own un-published conclusion, and write it in the article you've crossed a line that Wikipedia does not encompass. I am fine discussing this at greater length but would pose the following question: Why do you need to add a footnote to address matters that are supposedly "un-controversial"? Also for the sake of the discussion, it would be helpful to see some actual examples put forth of issues that have occurred where this particular footnote would have been helpful to finding an agreeable resolution and not antithetical to WP:NOR. Tmore3 (talk) 20:58, 20 December 2008 (UTC)

I think this entire conversation speaks to the problems that have arisen in porting what was originally a policy about research in a more or less scientific sense and secondarily a historical sense to a near-mirror of WP:V. The notion of "direct support" and what that entails is hopelessly vague, and the policy refuses to own up to this - most problematically, as I have said, at the primary sources section, but really throughout. This policy is astonishingly poorly written - the footnote helps inasmuch as the "deductive logic" approach is reasonable, but it seems to me to continue down the same failed path of carving out individual exceptions instead of addressing the root problems. I can think of few situations where this proposed change would be practically beneficial, because pure deductive logic is rarely used in actual argumentation and thought. Phil Sandifer (talk) 22:01, 20 December 2008 (UTC)

Tmore3 makes some excellent points, IMO. Phil Sandifer also points up some issues that are, I think, an entirely reasonable expression of what the state of affairs is on the wiki. As to Phil Sandifer's last sentence just above, I agree fully. The footnote currently in question has arguably succeeded in settling a question posed by users that are highly skilled in areas such as logic and mathematics. The rules of reasoning in such domains are in certain ways tighter and more focused than are the rules in other areas, such as, for example, politics, theology, art criticism, psychodynamic psychology, philosophy, history, sociology and a very wide array of other topic areas. The footnote I inserted, while imperfect from the perspective of technical logic, has drawn no criticism thus far from logical positivists or the math-interested. Fact is, I inserted the footnote in response to a thread that can be reviewed here. The basic issue had in a number of instances been subject to vehement complaints by users involved in topic areas where the word "synthesis" has a different connotation than it does in informal discourse. Mathematicians and formal logicians regularly synthesize in the normal course of "business", and it's an essential part of what mathematicians and formal logicians do. The accepted rules of synthesis are much more stringent in formal reasoning than is the case in informal language. Indeed it's sufficiently different that there's a whole very important topic called "formal language". This is why, at present, the footnote is there, because it's sufficiently important to acknowledge this issue w.r.t. these topic areas where the rules of reasoning are much more well defined than in most areas of study. Yet, this issue is, arguably at least, not generally relevant to most topic areas using our standard "informal language" in the context of deciding the frequently touchy issue of where "standard article writing in WP editors' own words" ends, and where "original synthesis" begins. ... Kenosis (talk) 05:13, 21 December 2008 (UTC)
I can appreciate the rationale for specific use within the subject areas outlined above by Kenosis. In principle however at least until the questions I raised earlier can be sufficiently addressed I remain unconvinced as to the urgency/frequency of this issue that it requires inserting further language into the document, even if it is in footnote form. I think if the footnote does stay the language defintely needs to be shored up somehow to avoid leakage elsewhere and be even more specific perhaps on the particular subject matter/scenarios where it might apply. Tmore3 (talk) 17:39, 21 December 2008 (UTC)

Please consider the following excerpt from the beginning of the article SKIP-BO,

SKIP-BO(/skɪpˈ-boʊ/) is a popular card game. In 1967, Ms. Hazel Bowman of Brownfield, Texas began producing a boxed edition of the game under the name Skip-bo. International Games, Inc. purchased the game in 1980. The company was subsequently bought by Mattel. It is a commercial version of the card-game "Spite and Malice". It includes 144 playing cards that have numbers on them ranging from 1 to 12 and 18 SKIP-BO cards totaling 162 cards. Alternatively, the 162 cards could be comprised of 3 regular decks of playing cards, including the jokers, with ace to queen corresponding to 1 to 12 and the kings and jokers corresponding to the SKIP-BO cards.

The last sentence is non-controversial and easily reducible to elementary deductive logic. However, it does not have a source, as far as I know. Thus it should be deleted according to WP:NOR. But if the new footnote were in WP:NOR, it would not be subject to deletion.

I don't think that the originators of NOR intended that it be used to delete sentences like this. I think they were concerned with material such as controversial unpublished physics theories and history analysis getting into the Wikipedia, rather than material that is non-controversial and easily reducible to elementary deductive logic. Also, we might consider whether a reader of that SKIP-BO article would prefer that the subject sentence remained in the article. Thank you. --Bob K31416 (talk) 18:05, 21 December 2008 (UTC)

Most certainly. This issue is fairly regularly debated on this page and elsewhere, sometimes heatedly. I think this kind of frequent disagreement is to be expected, because it is tied up with the sometimes-very-difficult question of where "article writing" or "original language" or "in one's own words" ends, and where "original research" begins. ... Kenosis (talk) 18:38, 21 December 2008 (UTC)
Again, that entire SKIP-BO article is OR, that game is unique to the American market, I can deduce that this article is in all likelihood factual, but that's not the point, and this is where I beg to differ (in respect to the accusation of me confusing OR issues with sourcing issues): we must assume the reader is absolutely ignorant of the subject matter, the means to verify the factuality of the content must be available directly to the reader, this is achieved by citing verifiable sources, its very plain and simple, and if you want a respected encyclopedia, longterm, this is the only way to deal with matters, right now different camps have different opinions on this but why have an NOR policy if it is consistently ignored in favor of NOR unless consensus agrees OR is OK? And why not state this explicitly instead of using all the pedantic idealized verbiage which nobody seems to care about unless a topic is controversial? Semitransgenic (talk) 19:34, 21 December 2008 (UTC)
Bob K31416's point was about the "A+B=C" scenario he was talking about in the last quoted sentence from the lead of the SKIP-BO article. The rest is a WP:V issue. For convenience, I'll simply copy and past a statement I recently made at WT:V: The rules, long established, are, essentially, "if it ain't cited, it can be removed, period." Such a removal, if there's any further disagreement, engages broader discussion process to (hopefully) arrive at a consensus. A user may, for instance, argue back at the person doing the removing that "no way, it's common everyday knowledge", or "it's already cited farther down in the article and WP:LEAD allows us to summarize it in the lead without citing it". Etc. etc. How this plays out depends on the material, how many are participating in editing the article, how controversial the material is, and so forth. ... Kenosis (talk) 19:56, 21 December 2008 (UTC)
[Please forgive me for inserting here, out of chronological order, some information that might clear up a possible misunderstanding of the SKIP-BO article regarding OR and WP:V. There is a source for almost all of the article in the external links at Rules for playing the Skip-bo game (PDF), except for the subject sentence that I mentioned above, which has no known source. Sorry for the interruption. --Bob K31416 (talk) 01:24, 22 December 2008 (UTC)]
more examples of sub-cultural nuances that are not encapsulated in the policy. My impression, based on somewhat limited experience across a narrow range of articles, is that insistence upon rigor seems to be almost frowned upon. The "if it ain't cited, it can be removed" is nice idea, but lets see how quickly my talk page fills up if I start clearing uncited material. Plus, it says "can be removed" meaning you first have to talk crap on a talk page about why you want to remove it, again the weight of consensus interfering with polices that are designed to improve the entire encyclopedia. I personally feel wikipedia currently weighs strongly in favor of inclusionism, the tendency to turn a blind eye to non controversial OR is quite strong. Or maybe I'm imagining this? But I don't think so, look at how convoluted the process to even propose an article for deletion is, it acts is a deterrent in my view. Semitransgenic (talk) 20:29, 21 December 2008 (UTC)
NB: "Sub-culture" is pretty much obsolete for some time now, having been largely abandoned in favor of the word "community". In this community, the decisionmaking is done by that ever-so-evasive "thing" we call "consensus". The primary exception is that User:Jimbo retains authority to dictate policy issues he or the Wikimedia Foundation deem adequately important to dictate. The informal version I gave of WP:V ("if it ain't cited, it can be deleted, period.") is a reasonable translation of a core policy set in place by Jimbo. The exact language can be found in the links given at WP:V. Having said that, ST, I do recognize the difficulties, including what you just said here. ... Kenosis (talk) 21:02, 21 December 2008 (UTC)
you may recognize them, but, and no offense intended here, you are still regurgitating the same old "community" dogma in defense of the glaringly obvious deficiencies that exist, and you are not alone in doing this. I'm trying to look at this using the perspective of someone who sees wikipedia consistently coming top ten on google and as someone who believes this really does mean that wikipedia has a responsibility to get things right, and get things right now, not whenever. There's a certain responsibility that comes with providing free knowledge on this scale, and the same standards should apply to all content, irrespective of consensus views on the matter. Semitransgenic (talk) 21:32, 21 December 2008 (UTC)
RE "wikipedia has a responsibility to get things right, and get things right now, not whenever." : That's a very tall order-- maybe WP can get a few billion of the TARP money and hire an editing staff.  ;-) (Speaking of which, the TARP article is presently a mess too. "Welcome to WIkipedia, the Volunteer Job that Never Ends" ;-) ... Kenosis (talk) 21:39, 21 December 2008 (UTC)
perhaps it is a tall order, though I was alluding to the notion of eventualism. Also, a deeply ingrained laissez faire mentality that exists across large sections of the community does not help matters, but that's just my opinion. Bots could probably look after a lot of the policy related editing matters that the humans on wikipedia are failing to address ; ) it really wouldn't be that difficult to do. Semitransgenic (talk) 22:41, 21 December 2008 (UTC)
LOL. Sure, a bot with an order, say as follows: "1) Eviscerate all content in excess of X characters not followed by [define 'inline citation']. 2) Eviscerate all Users with [define 'main namespace content-to-citation ratio'] equal to or greater than 1000-1" ;) ... Kenosis (talk) 23:46, 21 December 2008 (UTC)
Start er' up!! Seriously though I agree with much of what's been said as of late. Many of us know there are numerous articles with un-sourced content that go unchallenged for a period time for a multitude of reasons. While we can't do everything as quickly as we'd like, engaged editors, empowered by these core policies (despite all their supposed inadequacies) to challenge for accuracy has greatly improved this project as a whole. The major problem I have with arguments that essentially boil down to "it's common knowledge" or "it's obvious" is that both are extremely relative terms especially for an incredibly large and diverse audience. You don't have to present your educational credentials in order to read an article involving math or edit it for that matter, and it is dangerous if not wholly unfair to the anonymous reader to make an assumption that "everyone should know this and be able to deduce this" therefore I don't need to worry about providing a source to verify it. In regards to the original issue, I think it's usefulness is wholly unproved unless someone is willing to provide some convincing examples of where this footnote has stopped some actual re-occurring content disputes and is not in disagreement with current WP:V and WP: NOR policy as it reads currently. Tmore3 (talk) 01:05, 22 December 2008 (UTC)
I have no evidence in support of the idea that the footnote has stopped actual recurring content disputes. AFAIK, though, there've not been further complaints about the issue on this page, at least not from the perspective of those who do mathematical or logical synthesis in their areas of expertise. I sense that it's one of those situations where, given the wide range of material presented on the wiki, not everyone will ever be satisfied no matter which way this particular issue plays out. As to the presence of the footnote, I support it but I'm certainly not hellbent on keeping it. ... Kenosis (talk) 16:08, 22 December 2008 (UTC).
I am going to remove it for now as it has now twice been confused as policy for which there is no precedent currently in WP:V or WP: NOR that supports this statement as a content standard. I am fully open to continuing the discussion about it's usefulness in practice but either way an editing instruction like this should belong as part of a broader policy POV fully embraced by both WP:V and WP:NOR or as its own essay/guideline. Tmore3 (talk) 21:16, 22 December 2008 (UTC)
Please leave it off. If it's really "so obvious" with elementary logic, it's unlikely to be disputed anyway. I've been up to my shoulders in synth lately--much of it probably seemed "so obvious" to the editors putting in there. Professor marginalia (talk) 21:44, 22 December 2008 (UTC)
No problem by me. As is evident in the archived thread I linked to above, I was initially opposed to this sort of caveat and later became persuaded that such a caveat was reasonable as a footnote. Come to think of it, though, maybe next time a mathematician or whatever comes in here raising cain, Professor Marginalia can handle the talk thread and do the explaining. ;-) ... Kenosis (talk) 22:33, 22 December 2008 (UTC)
Professor marginalia, Thank you for weighing in on this discussion. The more the merrier. I'm very interested in the way people edit various situations in articles as a practical matter. Regarding your comment, "If it's really 'so obvious' with elementary logic, it's unlikely to be disputed anyway", if you came across a statement in an article that you found obviously true and that you had no objections to, except that you also found that it obviously ran afoul of WP:NOR, would you delete it? Thanks. --Bob K31416 (talk) 22:47, 22 December 2008 (UTC)
I don't think I would delete what I consider "obviously true" content just because it isn't sourced, no. Not for that reason alone. Professor marginalia (talk) 23:51, 22 December 2008 (UTC)
Surely it would be appropriate to tag rather than delete in this instance? Semitransgenic (talk) 00:10, 23 December 2008 (UTC)
Professor marginalia, If I may follow up, let's consider the excerpt from SKIP-BO that I gave previously. Here's the relevant part,
...It includes 144 playing cards that have numbers on them ranging from 1 to 12 and 18 SKIP-BO cards totaling 162 cards. Alternatively, the 162 cards could be comprised of 3 regular decks of playing cards, including the jokers, with ace to queen corresponding to 1 to 12 and the kings and jokers corresponding to the SKIP-BO cards.
The first sentence can be sourced. The second sentence cannot be sourced, as far as I know. It is true but not obviously true since you need to do some arithmetic, etc. to see that it is true. The second sentence is OR that runs afoul of WP:NOR. Would you delete it? Thank you. --Bob K31416 (talk) 00:50, 23 December 2008 (UTC)
Yes, I would but not because it uses arithmetic. Skip-Bo is a trademarked game, with its own set of cards, and the fact that one could play a similar game with an alternative set of cards is irrelevant. It's not encyclopedic to include hypothetical scenarios. What's more encyclopedic is to try and find sources that describe the ancestry or derivatives of the game itself. Skip-Bo is very possibly based on a traditional card game. Professor marginalia (talk) 01:06, 23 December 2008 (UTC)
Unfortunately, when using actual examples other issues can come up, like the one you mentioned above, which are a digression from the issue regarding the OR aspect of the sentence in question. Anyhow, thank you for your response. --Bob K31416 (talk) 13:20, 23 December 2008 (UTC)

<unindent>The footnote should perhaps have used a word like "uncontested" instead of, or in addition to, "uncontroversial". This issue appears to me to be one of those little glitches in policy logic that'll perhaps never get settled completely. ... Kenosis (talk) 23:38, 22 December 2008 (UTC)

This is the kind of problem that I worry we're inviting even more of with the "so obvious" language. Why source claims when it's "so obvious" that any child can see the association to "Tarde, Le Bon and Locke". Professor marginalia (talk) 03:45, 23 December 2008 (UTC)
Sheese. A rather inane and angry rant by our anon-IP friend on that page. FWIW, you can quote me on that, though it appears to me your response to that user was direct and quite adequate. ... Kenosis (talk) 04:52, 23 December 2008 (UTC)
I think we have to be careful not to use any unpleasant experience as a reason not to change WP:NOR.
The requirement of "non-controversial" or "uncontested" would not engender these unpleasant experiences because it is such a tough requirement. For example, if someone disputes whether a statement is "easily reducible to elementary deductive logic", then the situation has become controversial and contested. The disputing editor is in control of the situation. He can delete the subject statement. This is an important safeguard that should prevent the misuse of footnote 2. --Bob K31416 (talk) 13:20, 23 December 2008 (UTC)
who wants to waste even more time debating such matters on on a talk page? I don't agree that this footnote will benefit us. Semitransgenic (talk) 15:04, 23 December 2008 (UTC)
I mean on an article talk page where such a contest might arise, not this talk page. Semitransgenic (talk) 15:09, 23 December 2008 (UTC)
Thank you for your comment but the present topic was the "non-controversial" or "uncontested" requirement of footnote 2. Would you care to give your opinion about whether or not you think that requirement is an adequate safeguard against the abuse of footnote 2? We can discuss other issues regarding footnote 2 later but for the sake of an orderly discussion, lets stick to one topic at a time. Thank you. --Bob K31416 (talk) 17:04, 23 December 2008 (UTC)
I'm sticking to the topic, in summary, you want to graft the footnote from syn and apply it to OR in general, I disagree with this measure and choose not to support such a change. There can be no abuse of the footnote as long as the context of its use is specific to acts of synthesis. Semitransgenic (talk) 17:16, 23 December 2008 (UTC)
Perhaps it's just a misunderstanding but I think we were currently just discussing footnote 2, which I mentioned 3 times in my last message. From your last sentence, "There can be no abuse of the footnote as long as the context of its use is specific to acts of synthesis." Then, does this mean that you wouldn't object to having footnote 2 in WP:NOR? Thank you. --Bob K31416 (talk) 17:39, 23 December 2008 (UTC)
Let's be specific so everyone is clear as to what is being discussed... Bob, are you asking whether the footnote that is currently attached to the WP:SYNT section should remain? (I would agree to that) Or are you asking whether a similar footnote should be placed elsewhere in the policy? (I would probably not agree to that) Blueboar (talk) 17:48, 23 December 2008 (UTC)
Blueboar, Thank you for asking. Perhaps other people too who are following this do not realize that footnote 2 has been deleted. The current discussion is regarding footnote 2, not the footnote that I proposed. And thank you for your opinion that footnote 2 should be part of WP:SYNT which is a section of WP:NOR.
Here's the deleted footnote 2 for reference:
The rule against "A and B therefore C" does not, in general, refer to statements A,B and C that are non-controversial and easily reducible to elementary deductive logic." See also, WP:Verifiability#Burden_of_evidence
--Bob K31416 (talk) 17:59, 23 December 2008 (UTC)
Now getting back to the "non-controversial" or "uncontested" requirement of footnote 2. As I mentioned above, this requirement should be a safeguard against abuse of footnote 2. (I put that message in bold type so that it is easier to locate.) Semitransgenic seemed to agree with this for footnote 2 when it was in the context of WP:SYNT. With this line of discussion, I'm trying to address the concern regarding footnote 2 that it could be abused. Does anyone believe this safeguard is inadequate? If so I would be interested in reading your reasons why you think the "non-controversial" or "uncontested" requirement of footnote 2 isn't an adequate safeguard. Thank you. --Bob K31416 (talk) 18:21, 23 December 2008 (UTC)

(outdent)I actually missed where that item got deleted, so presumed it was in place, but I also seem to recall that when the footnote was originally proposed its intended application was for use in science and math based topics, however, I don't especially see that it can be abused if used in other contexts, because synthesis is synthesis, it's self-evident, this footnote offers an abuser nothing by way of circumventing the policy on synthesis. Semitransgenic (talk) 18:52, 23 December 2008 (UTC)

Bob, WP:SYN does not say with an A + B = C argument it's okay to come up with your own C if it isn't published so long as you can make a deduction from A & B. In fact it says completely the opposite. I appreciate the attempt to provide an example with the Skip-Bo article, essentially what was presented though is an article that uses an instruction manual as its only source and then advocating for a need to insert more unreferenced material of questionable encyclopedic worth. One of Wikipedia's goals as far as content is to aim towards more verifiability; why would you include and more importantly what need currently exists for a statement of ambiguous weight and scope that makes editors feel they can draw and publish as many un-sourced conclusions as they want just as long you don't get challenged. Tmore3 (talk) 22:59, 23 December 2008 (UTC)

Tmore3, Thank you for your response but I was looking for your opinion of whether the "non-controversial" or "uncontested" requirement is a safeguard against abuse of footnote 2. There are a number of aspects of footnote 2 that we can discuss but the aspect that is currently being discussed is whether or not the safeguard would work to prevent abuse. Perhaps your silence regarding this point suggests that you feel it would be an adequate safeguard against abuse? If not, please give your reasons why it isn't an adequate safeguard. Thank you. --Bob K31416 (talk) 00:47, 24 December 2008 (UTC)
Tmore3, I just reread your last message and part of your last sentence gave me an idea for a situation where footnote 2 wouldn't be helped by the safeguard. There may be a problem with an increase of unnoticed violations of WP:NOR. With footnote 2, some editors are bound to misinterpret the meaning of "easily reducible to elementary logic" and introduce some material that may continue in an article because it is not noticed. It is subject to deletion but it remains because it is unnoticed.
So it's a question of which is worse: a) the exclusion of worthwhile, informative, and credible material, that conforms to the requirements of footnote 2 but not WP:NOR as it stands currently without footnote 2 or b) the increase in questionable material as a result of misinterpreting footnote 2, but material that is subject to deletion if it is noticed. In other words, the questionable material is subject to deletion but it isn't deleted because it is unnoticed.
I think the suggestion "be bold" came about because people were afraid to contribute to Wikipedia because they might run afoul of a rule. I think that footnote 2 would result in an increase in the number of worthwhile contributions to the Wikipedia, but I have to admit that it's not clear how much this would be offset by objectionable material that might come into the Wikipedia too. I think this probably ends the discussion for me. Happy holidays everyone. --Bob K31416 (talk) 18:23, 24 December 2008 (UTC)
(My following message is a belated addition to my message above and is out of chronological order. Also, I hope my style of bringing up arguments for both sides of an issue isn't too confusing. I'm just trying to follow the truth wherever it may lead.)
After sleeping on it, the effect of unnoticed violations does not appear to be a good reason for excluding footnote 2. If a misinterpretation of footnote 2 results in an OR violation in an article, the violation is still subject to deletion and I expect that it will eventually be deleted. On the other hand, if worthwhile, informative, and credible material is included in an article, but wouldn't have been allowed without footnote2, then it is not subject to deletion because of NOR. If not for footnote 2 the worthwhile material would be subject to removal and is not secure, as far as WP:NOR is concerned. Without footnote 2, any editor who feels it is his duty to remove violations, no matter how worthwhile the contribution, will do so. With footnote 2, such an editor is not compelled to remove the worthwhile contribution. --Bob K31416 (talk) 17:07, 25 December 2008 (UTC)
the unnoticed situation is what we currently have, because consensus overrules policy. OR stays, unless someone makes a fuss, if they make a fuss, they are then forced to engage in a hoop jumping expedition to deal with it, the footnote will simply exacerbate an already tiresome situation. Semitransgenic (talk) 18:34, 24 December 2008 (UTC)
With the "non-controversial" requirement of Footnote 2 hoop jumping isn't exacerbated. Simply objecting is enough justification to delete because it is then a controversial situation. --Bob K31416 (talk) 18:17, 25 December 2008 (UTC)
Bob brought up a question about this on my talk page, with another long-time editor jumping in and commenting about the issue here. The more I think about this issue, the more I'm inclined to advocate a bit more strongly in support of an explicit caveat of this kind. In my observation it is acceptable to remove tendentious or spurious demands for citation where no genuine controversy exists that source A+B really are saying "C"-- indeed that is standard editing practice (one aspect of writing in one' own words rather than quoting the sources verbatim). As well, it's common practice to make simple deductions, for example, of the kind given in the SKIP-BO example and of the kind we very commonly find in articles using demonstrations of mathematical formulas or formal logic. The footnote presently under discussion gives an explicit policy basis for this. I'm not even sure the extra word "uncontested" is necessary. The difference between "description" and "interpretation" presently under discussion a couple sections below only partially addresses the issue from a policy standpoint. Another way of thinking about it is to differentiate between, on the one hand, uncontroversial synthesis of the kind that constitutes plain article writing in one's own words rather than in direct quotes, and on the other hand "original synthesis" or "synthesis serving to advance a position". It's the latter that is the reason for WP:NOR and WP:SYN. ... Kenosis (talk) 19:11, 24 December 2008 (UTC)
I agree with Kenosis. Footnote 2 would be beneficial to the Wikipedia. Furthermore, I don't think that the originators of WP:NOR ever intended that it be used to exclude worthwhile material that is credible and doesn't involve controversy. For example, here's an excerpt from one of the messages of Jimmy Wales,
"The basic concept is as follows: it can be quite difficult for us to make any valid judgment as to whether a particular thing is _true_ or not. It isn't appropriate for us to try to determine whether someone's novel theory of physics is valid, we aren't really equipped to do that."
Footnote 2 admits material involving cases where it is easy to make a valid judgement whether a particular thing is true or not. Jimmy Wales intention with WP:NOR was to exclude material where it is difficult to determine if something is true, not material where it is easy to determine if it is true, such as the cases covered by footnote 2. If a question of whether or not it is easy to determine if it is true arises, then the material is subject to deletion according to footnote 2 because it has become controversial. --Bob K31416 (talk) 18:34, 25 December 2008 (UTC)
Perhaps needless to say, I agree. I think that the presently removed footnote took spurious or tendentious tagging of non-controversial syntheses out of the mix, so to speak, by setting forth a straightforward statement of this particular issue, but without unnecessarily clogging up the body text with peripheral caveats. ... Kenosis (talk) 06:08, 27 December 2008 (UTC)

The distinction and clarification in the now-removed footnote is valid, but should be in the text not as a footnote. One puppy's opinion. KillerChihuahua?!? 14:51, 28 December 2008 (UTC)

Adding: I suggest a logical place to incorporate this would be the last para of the lead: "The "Original research" rule does not forbid routine calculations (like adding numbers, rounding them, turning them into percentages, putting then on a graph, or calculating a person's age, or giving the distances between geographical points) that add no new information not already present." as it would be a logical expansion of that. Concur, disagree, comment? KillerChihuahua?!? 14:55, 28 December 2008 (UTC)
For what it's worth, I would be much more inclined to support something akin to KillerChihahua's proposal as worded above as it makes two important distinctions from what was discussed earlier; one, it narrows application to routine mathematical calculations (which is why the footnote was proposed in the first place to address math and science subjct matter) and as stated earlier, a caveat is a caveat and should not be buried away in the bottom of the document. I think this proposal would be best served with its own discussion topic since we are now talking about a clear addition to the policy body of WP:NOR and frankly this one has become quite unwieldy. Tmore3 (talk) 17:43, 28 December 2008 (UTC)
Thank you. I too think the recently added last paragraph of the lede, as suggested by KillerChihuahua, will go a long way towards solving this issue without diminishing the basic principles of WP:NOR. I thank VasileGaburici for raising it several months ago, and everybody else involved in working towards what appears to be a sensible solution. ... Kenosis (talk) 04:44, 31 December 2008 (UTC)
I think there's a misunderstanding here. In KillerChihuahua's message, the quote regarding calculations is already in WP:NOR. I think Tmore3 wasn't aware of that. Didn't KC suggest incorporating the material from the deleted footnote 2 into the text of the paragraph about calculations? --Bob K31416 (talk) 06:27, 31 December 2008 (UTC)
I think KC does advocate that, though the brief note about routine calculations ought, I would hope, satisfy the issue brought up several months ago by VasileGaburici here. And just to help clarify, the current last paragraph of the lede was added here and in the two edits immediately thereafter. ... Kenosis (talk) 22:37, 31 December 2008 (UTC)

Forest for the trees

OK - a lot of people are getting hung up on the Derrida and Searle example. When this is combined with a rather bad failure to understand this example (for instance, calling Derrida and Searle's arguments unpublished), let's move to a general hypothetical to see where this becomes so toxic.

Imagine Person A, a researcher working in a technical, specialist area. His work is the subject of multiple independent reliable sources, and thus we have an article, Person A, that, among other things, describes his work.

Person B publishes, in a peer-reviewed journal, a critique of Person A.

Person B's critique is a reliable secondary source, and it is wholly appropriate to include it in the article. Whether or not there are any secondary sources on Person B's critique, it is, in and of itself, a significant point of view. NPOV mandates its inclusion. In it goes. And as a secondary source, it can be summarized freely, without concern about specialist knowledge.

Person A responds to Person B's critique - in an interview, in a peer-reviewed article, on his blog - this doesn't matter. All that matters is that Person A responds, as Person A's self-defense of his work is self-evidently a significant viewpoint on the subject. But as Person A's response is a primary source for the purposes of Person A, the standards for inclusion are far harsher - large portions of the technical work cannot be summarized, as the summaries would not be clear to a lay audience.

This presents a NPOV problem - Person B's criticisms of Person A are allowed far more latitude than Person A's defense, and can be described in thorough detail, while Person A's response cannot be nearly so detailed or substantial.

This is a problem for every single article about someone who works on a specialist topic and has been the subject of any criticism. Which is to say, for every single article about someone who works on a specialist topic. Including numerous BLPs, which we are overtly allowing to be slanted towards criticism.

This is a hole that has to be plugged. This is not optional - WP:NOR cannot set up a policy that undermines NPOV on a massive host of articles. It is simply not permissable - WP:NPOV is the prevailing policy in all disputes with other content policies, and other policies must be revised to match it.

Note that there is no guarantee that secondary sources will patch this. When you get out to specialists who are less massively notable, the secondary sources get thinner quickly. And so the leg up that criticism gets becomes more and more pronounced. Phil Sandifer (talk) 22:37, 28 December 2008 (UTC)

I see a few flaws in your argument... when you say: Person A responds to Person B's critique - in an interview, in a peer-reviewed article, on his blog - this doesn't matter. All that matters is that Person A responds, as Person A's self-defense of his work is self-evidently a significant viewpoint on the subject. But as Person A's response is a primary source for the purposes of Person A, the standards for inclusion are far harsher - large portions of the technical work cannot be summarized, as the summaries would not be clear to a lay audience.
You make an assumption that I don't think is valid... I think it is always possible to write a summary that would be clear to a lay audientce. It may not be easy, and YOU might not be able to do it... but someone ELSE could.
Then you say: This presents a NPOV problem - Person B's criticisms of Person A are allowed far more latitude than Person A's defense, and can be described in thorough detail, while Person A's response cannot be nearly so detailed or substantial.
You have two flaws here, the first is related to the one I raised avove... I think you can summarize the response. But if you can not... then you have several options... you can quote Person A's response directly ... you can find secondary sources that comment on A's response. The second flaw is that you seem to think that NPOV requires viewpoints to be given equal time and space... it doesn't. This is covered by the Undue Weigh clause... Person B's criticisms might well deserve more weight than Person A's response.
In other words, I find your argument that this section of NOR is in conflict with NPOV to be flawed. There is no conflict. Blueboar (talk) 23:01, 28 December 2008 (UTC)
But if you can not... then you have several options... you can quote Person A's response directly ... you can find secondary sources that comment on A's response.
The problem is that you didn't need to do any of these things in order to use person B's critique. It's perfectly fine to just summarize B's criticism yourself. You may have "several options" to summarize A, but while these options are several, they're still nowhere near as good as or as straightforward as the one big option you're only allowed to use for B.
Moreover, you're confusing "comment" and "summary". There may be lots of secondary sources which comment on A's response, while still being few or none which summarize it. Ken Arromdee (talk) 00:41, 29 December 2008 (UTC)
Yes, exactly - all of these options still end up giving Person A a harder road into the discussion than Person B, which is particularly problematic on Person A. Phil Sandifer (talk) 01:05, 29 December 2008 (UTC)
Well, you'll forgive me if I'm hesitant to take the general case on your goodwill that the response can be summarized. I mean, I see some major problems - the policy can very, very easily be used abusively here. In fact, it takes some rather creative reading of the policy to avoid the abusive use.
Similarly, I think we run into a real problem with the idea that person A's response deserves less weight. Perhaps this is true for an article on the topic in general - if, for instance, Person A and B are disagreeing on Foobitzes, Foobitz may well deserve to have unequal weight. Then again, on Foobitz, both Person A and Person B's articles are secondary sources, so the problem doesn't exist. However for Person A, I have a real problem with the idea that Person A's viewpoints are less important than Person B's, and if Person A happens to be a BLP, I think we run into some very, very serious problems. Phil Sandifer (talk) 01:05, 29 December 2008 (UTC)
to a first approximation, person A is an expert on the intended meaning of his own theories. He is not a reliable expert on whether his theories or correct, or whether objections to them can be answered. If B objects, and A explains further, no matter where, his explanation can be used, with an indication of where it comes from. Neither A nor B are experts on whose interpretation is correct. In B says A's theories or wrong and his explanations are inadequate, no matter how expert B, that's his own opinion, and has to be cited as such, no matter where its published. The academic world originated and continues, essentially as a debating forum, and may reach consensus, but does not reach truth. DGG (talk) 02:54, 29 December 2008 (UTC)
I'm not sure, all told, how relevant that is, given that what we're dealing with here is less a matter of reliability than NPOV. If B is published in a peer-reviewed journal, it is a reliable source with a significant viewpoint on A. If A defends his theories, it is a significant viewpoint on A as well. Regardless of the expertise or correctness, both sides need to be presented for the purposes of a comprehensive and neutral view on A. Phil Sandifer (talk) 02:58, 29 December 2008 (UTC)
I think this comes back to a point I made earlier... attempting to write articles that are too specialized for Wikipedia. I don't see a need for our articles to be "comprehensive"... or at least, not to be so comprehensive that we can not rely on what secondary sources say on a topic. The goal of Wikipedia is give an overview of the topic... and that can be done by relying on secondary sources just as well as from summarizing primary sources (and if there are no secondary sources, I have to wonder whether the topic is notable in the first place)Blueboar (talk) 03:10, 29 December 2008 (UTC)
Well, it's pretty easy to find tons of topics that are extremely specialized but are still covered in two reliable, independent secondary sources, so they're going to pass Wikipedia:Notability pretty easily. And I don't think that a "remove specialist coverage" approach is going to fly. Phil Sandifer (talk) 03:30, 29 December 2008 (UTC)
Yes. It's true that the Derrida example is problematic support of a policy wording change since in that case he's accused of "obscurantism". And it's being argued here he should be used as a primary even though doing so would require "specialist knowledge" for the interpretation. If his words are presented to us as "inadequately" described in secondary sources, why give any editor's word more credence? So I think Derrida is relevant to the question. It emphasizes the problem of wikipedians coming to novel interpretations from those found in the secondary sources. But I'm not sure I understand why it's presented as okay to offer novel "interpretations" of the secondary sources either--it isn't. Professor marginalia (talk) 03:19, 29 December 2008 (UTC)
So is your proposal to bar all use of sources in ways that require specialist knowledge? Phil Sandifer (talk) 03:30, 29 December 2008 (UTC)
For a novel interpretation, yes. I'm not proposing to bar such use of sources. I claim this is the way it is now and it should stay that way. Professor marginalia (talk) 04:27, 29 December 2008 (UTC)
And how do you propose to tell a novel interpretation of a secondary source from an unoriginal one? Phil Sandifer (talk) 04:31, 29 December 2008 (UTC)
Are we getting sucked into absurd abstractionism? This is easily remedied for a well covered subject such as the Derrida example offered here. If we're to dispense with it, we need examples. Off the top of my head I can think of 4 scenarios that I think demonstrate WP adequately covers all the bases. If a claim taken from a source is an interpretation that requires "specialist knowledge" to properly interpret, then it is either a) original research, b) a very fringe-y claim, c) not noteworthy enough to mention at wp or d) an empty excuse offered to avoid the work of finding existing reliable sources. There may be others, and if so, I'm open to contrary examples to illustrate problems that I haven't seen yet. Do you have any? Professor marginalia (talk) 04:47, 29 December 2008 (UTC)
The Derrida example I started this section to get away from? How helpful. Your proposal appears to amount to "no specialist topics whatsoever on Wikipedia." Good luck getting that through. Meanwhile, let's work on the project we actually have. Phil Sandifer (talk) 04:52, 29 December 2008 (UTC)
You offered it as an example, and pulled it back. I acknowledge that. I'm asking for another one. Can you give one? Professor marginalia (talk) 04:56, 29 December 2008 (UTC)
Given two sections down, along with a general pointing at a few dozen more. Phil Sandifer (talk) 05:14, 29 December 2008 (UTC)
And I don't assert there be "no specialist topics whatsoever". I assert there be no "unsourced, specialist-only interpretations of topics whatsoever". Professor marginalia (talk) 04:58, 29 December 2008 (UTC)
We're not talking about unsourced material. We're talking about material that is sourced to difficult, specialist works. Phil Sandifer (talk) 05:14, 29 December 2008 (UTC)
"unsourced interpretations"--if the "interpretation" of any source is unclear except to a wikipedian self-proclaiming to have "specialist" skills, it is it OR? Yes. Professor marginalia (talk) 05:31, 29 December 2008 (UTC)
And again, the question becomes, how do you propose to deal with this? We can not allow description of sources that are difficult for non-specialists to understand. Or we can... what, exactly? Phil Sandifer (talk) 05:36, 29 December 2008 (UTC)
Descriptions are okay, for primary or secondary sources. It's the problem of overlaying on top of them some novel interpretation or analysis or conclusion, etcetera from those sources which we're focusing on. Professor marginalia (talk) 05:50, 29 December 2008 (UTC)
Putting aside my dislike of the description/interpretation distinction, I should note that what we are talking about here is not novel interpretations of sources, but rather descriptions of the arguments - very, very basic readings. It's just that the sources involved are difficult and rely on technical knowledge such that not just anyone can perform even a basic reading. But it is not as though we are choosing among multiple competing interpretations of a text - it is that we are dealing with texts that some people will understand and be able to describe on a basic level, and others will have no understanding of whatsoever. Phil Sandifer (talk) 13:31, 29 December 2008 (UTC)
In real life there are hundreds of thousands of reports publiched every year on lab experiments. Most of them are not encyclopedic. If they are important enough for Wikipedia they will be replicated by other scientists and thus there will be numerous citations. The most useful citations are the ones that (in their opening pages) also summarize or review the consensus of scientists on the issue. It's the consensus we want to report (or the existence of several live competing theories). Rjensen (talk) 10:17, 31 December 2008 (UTC)


Good faith

If I may make a comment, accusations of bad faith and declaring that my discussion is tendentious do little to make me think that the people I have spent quite a bit of time and effort engaging in dialogue with are actually interested in anything other than shutting the discussion down and getting me to go away.

If this is the case, so be it. I will approach the matter differently, and with less regard for attempts at discussion, as clearly nobody is interested in discussion. However, I sincerely hope it is not the case. In any case, I remain very much interested in continuing this discussion and figuring out how to clarify this policy so it does not suggest that NPOV violations are the way to go, and does not describe the process of reading in ways that are utterly bizarre to anyone who actually studies that process.

If you are not interested in that discussion, that's fine, but "I don't want to discuss it, so my preferred wording wins because it's already in place" is not an acceptable tact. Phil Sandifer (talk) 17:51, 31 December 2008 (UTC)


Phil, please don't take this approach. For my part, it isn't that "I don't want to discuss it"... it's that we have discussed it... over and over again. We are all beginning to repeat ourselves.
There comes a time when people get tired of repeating what has already been said. There comes a point where continuing to argue about it is unproductive and pointless. Blueboar (talk) 18:18, 31 December 2008 (UTC)
I would rather not take this approach, but when one editor barges on and starts openly accusing me of bad faith, well, I consider my options. Especially because, frankly, I dislike the idea that inertia can be used to bully a discussion shut, which is effectively what is happening here. It is not as though the discussion has been moving in circles. I'll agree there are some core issues that are not yet worked out. But I certainly don't feel like it's moving in circles, which makes the attempt to shut it down extremely distasteful. Phil Sandifer (talk) 22:41, 31 December 2008 (UTC)

An aside

At the risk of getting us obscessed with Derrida again... but I have been thinking about Phil's comments and have a question... we have been opperating on the assumption that Derrida's reply to Searle is a primary source. But is it? Can we not make the argument that it is a secondary source?

Start with Derrida's original work (I have lost track of the title of that work... but I am referring to whatever it was that Searle was critiquing ... I will call it X)... if we call that the primary source, then Searle's critique of X is a secondary source, I think we are all agreed on that. But isn't Derrida's reply also a secondary source in relation to X? Yes, it happens to be authored by the same person as X, but it is a seperate publication. In essence, it is Derrida's critique of X.

If I am not completely off base here, then doesn't this resolve at least part of the issue that Phil has been struggling with (at least it in the case of the Derrida article)... You can cite Derrida's reply as a secondary source for Derrida's own interpretation of his work.

I realize that we are trying to move the main conversation away from Derrida... which is why I am placing this in its own thread... but I do think it is worth considering. Blueboar (talk) 18:42, 29 December 2008 (UTC)

Derrida's reply to Searle can reasonably be taken as a secondary source both w.r.t. Searle's commentary and w.r.t. Derrida's work being commented upon. Searle's critique can be taken as a secondary source w.r.t. Derrida's work being commented upon. That is to say, they both explain, explicate, expand upon, or otherwise interpret the primary source(s), in this case Derrida's published work. ... Kenosis (talk) 01:34, 30 December 2008 (UTC)
And for the subject of Derrida himself? Phil Sandifer (talk) 01:44, 30 December 2008 (UTC)
Seriously? . . . OK. I think this is splitting hairs unnecessarily, but to the extent that Derrida's work occupies a significant portion of the article on Jacques Derrida, Derrida's commentary which explains aspects of his work can reasonably be regarded as a secondary source if used to elucidate one or more of Derrida's major works, and as a primary source if used to describe something about Derrida himself. Did you mean the latter? w.r.t. Derrida's life itself? say, something about his personality? IMO, that's splitting hairs too. If that's the focus, any autobiographical commentary, something written by Derrida being used to convey something about Derrida other than an explanation of some aspect of one of his major works, it would become a primary source. Consistently with WP:PSTS, a primary source should be used with the additional measure of care as indicated in that part of the policy. But I haven't read either Searle or Derrida's response to Searle, so I don't at the moment know what's in the Derrida/Searle debate and how it might be applied to giving WP readers a biographical perspective on Derrida himself. Fair enough?
..... Are there no secondary sources that discuss the Derrida/Searle debate? I should think there are plenty. If that's the case, WP:PSTS indicates that we should rely on the secondary sources to the best extent we can, rather than try to interpret the Derrida/Searle debate ourselves. The reason? Because to do that would be original research. ... Kenosis (talk) 03:21, 30 December 2008 (UTC)
Part of the problem here is that Derrida's response essay - Limited Inc - is probably better known than his original essay (in part because Limited Inc was the title used for the book both were republished in). So it's not like it's commentary that is separable from the rest of his work. But more significant here - as there are secondary sources on the Derrida/Searle debate - what do we do with debates between lesser known but still notable academics, where the sourcing is much thinner? Phil Sandifer (talk) 17:25, 30 December 2008 (UTC)
The Shusterman and Brennan example works also. In the article Brennan interprets Shusterman. Shusterman objects to interpretation made in Brennan. And Brennan objects to Shusterman's objection to his interpretation. I don't think the primary/secondary question is important here. What's important is how the source is characterized here. What's necessary at wikipedia is that the positions of both are described and attributed. What's not okay is for a wikipedian to present some position that isn't sourced, or present any interpretation, implication, conclusion, generalization, etc., of any position taken unless that interpretation, implication, conclusion or generalization itself is sourced. In the case of John v Mary, John's view is described and attributed, Mary's view is described and attributed to John's and Mary's own words where need be. But any further interpretation, such as "John's argument demolished Mary's position", or "Mary exposed John's errors" would need third-party sourcing. Professor marginalia (talk) 19:05, 29 December 2008 (UTC)
Though the more glaring problem with those statements is NPOV. But in any case, current policy causes problems even with description, as description that requires specialist knowledge to verify is explicitly forbidden. Phil Sandifer (talk) 21:14, 29 December 2008 (UTC)
Yes - both Derrida and Searle are secondary sources for speech act theory, which was the subject of their debate. The issue really springs up for the bios of the participants. If we want to specify that published writings of a person are not primary sources for articles about that person, or at least that published writings that do not make biographical claims are, that would clear up the matter to a great extent. Phil Sandifer (talk) 20:55, 29 December 2008 (UTC)
I think there may be some confusion between the concept of "first party vs. third party" sources, as opposed to the concept of "primary vs. secondary, vs. tertiary" sources?
Anything written by someone directly tied to the subject of an article (or the subject himself) would be a "first party" source for that article... but a first party source is not automatically a "primary" source... it could be a secondary or even a teriary source. Conversely, a third party source would be anything written by someone not closely with the subject of the article, but might be a primary source (or a secondary source, or even a tertiary source)
To show the difference... take a scholar who writes a standard text book for school children. The text book is a first party source in a bio article on that scholar, even though (being a standard text book) it would be considered a secondary (or possibly tertiary) source for its contents. Blueboar (talk) 21:47, 29 December 2008 (UTC)
While checking the Derrida/Searle sources and Shusterman/Brennan items I came across a letter in PMLA (Journal of the Modern Language Association) where a question is asked regarding 3 sentences attributed to Derrida. The explanation requires 9 paragraphs. The inquirer's view, someone concerned with "taking care of readers, making modification clear to them, making reference only to publicly accessible things, maintaining sentence coherence" relates to how we should be thinking here, whereas the respondent, in seeking to "[demonstrate], that none of these responsibilities can be taken for granted, and none of them is beyond philosophical scrutiny and critique" pretty much sums up Phil's position. For the purposes of writing an encyclopedia it's pretty clear which holds more value.
I think it's generally accepted that Derrida's writings are controversial. They have a reputation for being difficult, are written by a specialist, for consumption by a specialist audience, and, in general, it requires specialist training to decipher such writings. Also, Derrida's writings in English are translations of French texts, which complicates the matter still further: why not translate directly from the French texts if you are going to use a primary source? probably because this would require not just a Derrida specialist, but also a specialist in the use of the Frecnh language at an advanced academic level.
Derrida is not writing for the average reader, those analysing Derrida are not writing for an average audience, but, wikipedia is intended for general use by average readers, and should appeal to the widest possible audience, and it should not require that a reader have specialist knowlege of a topic. That does not mean to say specialist subject matter cannot feature: it should simply be in the right place, and presented in the right way. A biography article is not the place to explore highly academic minutae, and it is possible to write an adequate biog of any notable individual using secondary sources exclusively.
In writing an accessable article on a difficult subject, someone genuinely knowlegable of the subject area should know which writings are most suitable (if they put the interests of wikipedia first). Claiming that all secondary texts on a subject are inadequete is not a position that serves the community. If no secondary sources exist, notability may be an issue, but if it isn't, then the use of primary sources could be considered, but it's a matter of addressing specific cases as they arise, and in accordance with concensus. Removing the specialist knowlege not required provision is not a solution.
Regarding the Derrida/Searle argument Derrida, Searle, Contexts, Games, Riddles, Edmond Wright, Source: New Literary History, Vol. 13, No. 3, Theory: Parodies, Puzzles, Paradigms (Spring, 1982), pp. 463-477, appears to contain a useful and accessible overview.
Brennan and Shusterman, are written in relatively plain English, a million miles from Derrida. They are discussing rap/hip-hop, something any educated person with a command of the English language, and without specialist knowledge, can follow despite the usage of academic jargon; the same cannot always be said for Derrida.
I can post all of the above if anyone wants to examine the texts for themselves. Semitransgenic (talk) 22:41, 29 December 2008 (UTC)
I've got some real reservations about this - thought I have to look at the Wright article to comment further on it. When it comes to the Brennan/Shusterman debate, although some sections are easy, some are, to my mind, quite dense. Being deep in the academic bubble, I am perhaps not the best person to judge difficulty, but it is certainly my strong sense that passages of that debate would be exceedingly difficult. Phil Sandifer (talk) 23:50, 29 December 2008 (UTC)

Yup... I thought this might happen... we have once again slid off the track and are focusing on the details of Derrida again, instead of focusing on why we brought him up in the first place. To put the aside back on track... Phil, does distinguishing between a "first party source" and a "primary source" clarify any issues as they relate to the Derrida article (and if so, does it also help to clarify your issues with the primary source secotion in general)? If so, great. If not, why not? Blueboar (talk) 02:08, 30 December 2008 (UTC)

If we are willing to establish something like that a primary source for a biographical article would only be an autobiographical piece, not simply any piece of published writing, then yes, that would, I think, satisfy the bulk of my issues. Phil Sandifer (talk) 02:54, 30 December 2008 (UTC)
Unfortunately, that isn't accurate... and is getting the point of the distinction backwards. First/third party is defined by who writes the surce... primary/secondary/tertiary is defined by what is said in the source... whether the information or ideas that are contained in the source have been published before or not.
An autobiography is a first person source for an article on its author and is generally a primary source (the first publication of new information about the author), but the author could write a primary source on some other topic. In a bio article on the author, this source would also be a first party, primary source. In an about someone or something else, the first/third person designation would shift, but not the primary/secondary designation... it would become a third party, primary source. Blueboar (talk) 15:31, 30 December 2008 (UTC)
This is completely wrong - primary/secondary is not inherent to a source - a source can readily be a primary source for one topic and a secondary source for another. Phil Sandifer (talk) 15:39, 30 December 2008 (UTC)
Agreed... but primary/secondary is portable between articles that discuss the same topic... If a source is considered a primary source on a given topic, it continues to be a primary source on that topic no matter where you discuss the topic. In any case, I don't think your point changes how I am distinguishing between first/third party and primary/secondary (tertiary sources are similar to secondary in this reguard). Let me try lay out what I am talking about ...
  • First party, primary - Author is the subject of the article (or in non-bio directly connected to the subject), source is the first place of publication for ideas or facts relating to the topic under discussion.
  • First party, secondary - Author is the subject of the article (or in non-bio directly connected to the subject), source is not the first place of publication for ideas or facts relating to the topic of discussion.
  • Third party, primary - Author is not subject of the article (nor directly connected to the subject), source is the first place of publication for ideas or facts relating to the topic.
  • Third party, secondary - Author is not the subject of the article (nor directly connected to the subject), source is not the first place of publication for ideas or facts relating to the topic.
In other words, we are distiguishing between the author and the content of a source. The author determines first/third person... the content determines primary/secondary. Does this make sense? Blueboar (talk) 16:12, 30 December 2008 (UTC)
Hm. I'm a lot more hesitant of this reasoning - I think it broadens primary sources considerably from where we have them now. Especially when you get to "source is the first publication for ideas or facts relating to the topic," you make, at least in the humanities, the vast majority of sources into primary sources, as generally any published scholarship is presenting new ideas - that's the whole point of scholarly publication. Phil Sandifer (talk) 16:27, 30 December 2008 (UTC)
I certainly did not intend to broaden the definition of primary source. I was trying to summarize the definition that is used at the Primary source article into one simple sentence. Obviously, I did not do a good job of that. I am still at the conceptual stage, and not really focussing on exact language. I think the concept of distinguishing between author and content in this way makes sense. I think it may help resolve some of the problems people have been having with the PSTS section. If people agree with the concept, then we can work on language. Blueboar (talk) 17:02, 30 December 2008 (UTC)
Fair enough. For me, and especially based on PBS's observation above, the important place to close this is on the matter of outright technical specialists. Which makes me wonder if the better fix than this two-tiered system (which I'm having trouble wrapping my head around, despite the fact that I teach this stuff on the college level) would be some sort of acknowledgement of reliableness as a counterweight to primary sources. I'm not sure that something published in a peer-reviewed journal or by an academic press should ever, for our purposes, be considered a primary source. In the sciences, it seems like the policy is already implicitly making such a divide - the list of examples of primary sources gives "written or recorded notes of laboratory and field experiments or observations," which is very different from published scientific papers. Similarly, the difference here might be between a scholar's notes, journal, diary, etc and his publications. Which, I think, is a move that is not strained in terms of NOR - dealing with sources that come from the highest echelons of reliable sources in and of itself mitigates heavily against OR. Phil Sandifer (talk) 17:22, 30 December 2008 (UTC)
From the POV of an historian, I would definitely agree that articles that are published in peer reviewed journals are not normally considered primary sources (in history, primary sources are the documents cited in such journal articles: diaries, letters, newspaper accounts, monastic annals, etc., from the time we are writing about). That said, there is a bit of a grey area with history in that, over time something that would have been considered a secondary source when it was written may end up becoming a primary source today. A good example of this are medieval histories and annals... at the time they were written they would have been classified as secondary sources (relying on even older documents), but over time they shift designation and are, today, generally considered primary sources. Another factor historians have to consider is what era we are writing about... Take a scholarly paper on some aspect Saxon England, published in an accademic journal during the Victorian era. If we, today, are also writing about Saxon England, that article can be considered a secondary source (allbeit probably an out dated secondary source)... but, if we are writing about Victorian England it might be considered a primary source. Not easy.
This is one reason why, even though I agree with the principles layed out in the PSTS section, I too have problems with its language. I agree that we need to be extra careful when using primary sources (because it is easy to misuse them in ways that introduce OR). But that still begs the question of whether a given text actually is a primary source or not. I have long felt that we should warn of the OR dangers of using primary sources, but not attempt to define the term (instead, acknowledge that different accademic diciplines define the term differently, so we can not.)
Unfortunately, that is a battle that has been fought and lost. My views on the subject are in the minority and there is a clear consensus for the current language. Perhaps, with time, this consensus will change... but, at the moment, I have to accept this consensus and move on if I want to continue to contribute to Wikipedia. Blueboar (talk) 18:38, 30 December 2008 (UTC)
Well, though that Victorian example can probably be worked around via some hedge about "scholarly sources being used as sources on the subject of the scholarship." I continue to like this the more I look at it - I see little evidence that the primary source language was written to minimize the use of academic sources in any way, or that any consensus to do so exists. Phil Sandifer (talk) 23:31, 30 December 2008 (UTC)
The more we approach the problem as if disputes result from "Article 2, section 3.14" type loopholes or ambiguities, the more we lose sight of the issue the policy is trying to resolve. The problem is that at WP a primary source for X cannot be used as a source for any opinion about X. It's that simple. A primary source for X can be sufficient reference for a summary description of X, but with caution. One of the cautions about how to stay within the boundaries of "basic description" and away from "opinion about what X means" is that we limit "basic description" to that which is easily verifiable to any reasonable, educated person and doesn't demand "specialist knowledge" to evaluate. This clause focuses on the problem of using "opinion of X" for "articles about X" because it is so tempting and would be so easy, when writing about X, for editors to apply their own heuristic-type analysis (original research) of primary sources on X. This does not mean it's permissible to write "opinions of secondary sourced opinion of X" instead of "describing the opinions about X given in secondary sources". So there is no NPOV issue. The more we try to define the problem as if it were solvable by using definitions based on some Platonian Universal idea of primary or secondary source, the further afield we get from the real issue. If this clause somehow implies it is permissible to give basic descriptions of secondarily sourced opinion of X that are "not easily verifiable to any reasonable, educated person but instead require specialist knowledge to evaluate"---well that's the problem then with the wording. The description at WP of what a 2ndary source wrote must be easily verifiable as well. Professor marginalia (talk) 18:46, 30 December 2008 (UTC)
So you are suggesting that "All original descriptive statements about a source (regardless of type) must be easily verifiable by any reasonable, eductated person without specialist knowledge". I can fully agree with that. Blueboar (talk) 19:30, 30 December 2008 (UTC)
Yes. The policy now reads, "Drawing conclusions not evident in the reference is original research regardless of the type of source". To my way of thinking, if some secondary source is impenetrable to reasonable, educated persons and decipherable only to those with "specialist knowledge", then we would need tertiary sources to verify any reading of the secondary. It's a hard case to make, the way I see it, that we could have confidence that an impenetrable 2ndary source was used properly in a cite unless we consulted tertiary sources. Professor marginalia (talk) 20:09, 30 December 2008 (UTC)
I think that rules out huge swaths of advanced and specialized knowledge, and does a grave disservice to our mission. I also think that this only aggravates the NPOV problem - NPOV, as it stands, calls for all viewpoints that have been published in reliable sources. To exclude those that are specialist only does further damage to NPOV. Phil Sandifer (talk) 20:19, 30 December 2008 (UTC)
Do you have examples of the "huge swaths" of secondary sources relied upon at WP that aren't verifiable except to those with specialized knowledge? The argument seems a bit strained to me. Professor marginalia (talk) 20:36, 30 December 2008 (UTC)
Manuals and documentation for almost any topic along the lines of computer programming. Most research published in scientific journals. Large amounts of research published in humanities journal. Phil Sandifer (talk) 21:06, 30 December 2008 (UTC)
I beg to differ. Most reasonable, educated people with access to a source used have the ability to verify claims cited to these sources. You don't have to have specialist knowledge to do this. Professor marginalia (talk) 21:19, 30 December 2008 (UTC)
I like to think that I am both reasonable and educated, but when I flip through the papers from chemistry journals that my wife prints out periodically, I understand virtually nothing, and would be hard-pressed to verify summaries. Her experience with journal articles from my field is not dissimilar. Phil Sandifer (talk) 21:33, 30 December 2008 (UTC)
Again, this has nothing to do with whether the source is specialist in nature... It has to do with what we write and whether what we write can be verified by a non-specialist. If we write "Steven Humpledumpher's theory of Quantum Flexology postulates that the interconductive reeblefrob must interpunctuate with the bicammeral dorfnitz<cite to Quantum Flexology by S. Humpledumpher>" (a simple descriptive statement of Humpledumper's theory) then our readers need to be able to verify that this is what Stephen Humpledumpher actually said. They don't have understand what a reeblefrob or a dorfnitz is, nor how they could possibly interpunctuate... in fact they don't have to understand any of it... they simply have to be able to verify that the source says what we say it does. This is why page specific citations are preferable to broad citations to entire books... they help the reader find where in the book we got our information. Blueboar (talk) 20:45, 30 December 2008 (UTC)
I suppose my issue here is that I disagree with you on the possibility of this - I just don't think that a summary - even a good summary - is necessarily easily verified by someone who does not understand the source. Phil Sandifer (talk) 21:06, 30 December 2008 (UTC)
Actually, I think our disagreements may be based upon something more basic than that... I am not even sure that we agree on what a descriptive summary consists of. Blueboar (talk) 21:18, 30 December 2008 (UTC)
"I just don't think that a summary - even a good summary - is necessarily easily verified by someone who does not understand the source. I don't know where you get this idea. I've yet to see a problem with this. When we speak of "understanding the source", you simply have to have the ability to read what it says against what the claim attributed to it says. Professor marginalia (talk) 21:31, 30 December 2008 (UTC)-
Well, yes. The issue is that if I do not understand what the source says, I am hard-pressed to say that a given summary is accurate. Taking a random jargony sentence from a science article: "The epithelial amiloride-sensitive sodium channel constitutes the rate limiting step for sodium reabsorbtion by the epithelia lining the distal part of the kidney tubule, the urinary bladder and the distal colon." I don't know what that means. Short of a direct or virtually direct quote, I do not know how I would verify that. If "epithelial amiloride-sensitive sodium channel," "rate limiting step," "sodium reabsorbtion," "epithelia lining," "distal part," "kidney tubule," or "distal colon" do not appear, word for word, in the summary, I am hosed, because I do not know any synonyms for any of those terms. Phil Sandifer (talk) 22:19, 30 December 2008 (UTC)
That passage leaves very little room for appropriate synonyms, so you wouldn't have much trouble. No claim taken from it at WP could deviate far and still be true to what it really says. Seriously. Those are almost all proper terms with a specific meaning. Professor marginalia (talk) 22:57, 30 December 2008 (UTC)
That's all well and good - but I do not know that, and thus do not feel confident in verifying a summary. For instance, for our purposes, if it is a valid summary, combining the list of three body parts at the end - which all appear to be part of the digestive system - into a statement about how it is a limiting step for sodium reabsorbtion by the digestive system. If that's a valid move, it's a preferable summary for our purposes, because it is shorter and somewhat more direct - but I have no idea if it's a valid summary. In another example: "The application of a chiral crown ether to the separation of optical isomers in capillary zone electrophoresis is described for the first time." I can tell, looking at the rest of that abstract, that we're dealing with pharmeceutical research in some fashion. I know that capillaries are part of the circulatory system. Does that mean that a summary that says that "chiral crown ethers have applications in the separation of optical isomers in the circulatory system" is accurate? I wouldn't know, but if it is, that would be a preferable summary for at least some purposes to retaining "capillary zone electrophoresis." A second problem comes in dropping adjectives - to reduce clutter and tighten language, it might be desirable to cut an adjective like "chiral," which, in many contexts, would be a perfectly fine move to summarize. Would such a summary change the meaning? I wouldn't know. Phil Sandifer (talk) 23:25, 30 December 2008 (UTC)
It sounds as if maybe you haven't read the guidelines about using scientific journals. I don't have the wl at my fingertips or I'd use them here. What you're trying to do is apply a generalization from a specific claim in a primary source document, and you won't find good editors who have specialized knowledge doing it like this. You will find that in scientific articles on WP that you won't find much liberty taken in claims cited to primary source documents in the sciences, for good reason. I'm rushing this comment, but will try to explain better when I get back. Professor marginalia (talk) 23:52, 30 December 2008 (UTC)
Here's the guideline I was thinking of: Wikipedia:Reliable_sources_(medicine-related_articles). As it says, "In general, Wikipedia's medical articles should be based upon published, reliable secondary sources whenever possible. Reliable primary sources can add greatly to a medical article, but must be used with care because of the potential for misuse. For that reason, edits that rely on primary sources should only describe the conclusions of the source, and should describe these findings clearly so the edit can be checked by editors with no specialist knowledge. In particular, this description should follow closely to the interpretation of the data given by the authors, or by other reliable secondary sources. Primary sources should not be cited in support of a conclusion that is not clearly made by the authors or by reliable secondary sources, as defined above." So I can understand why you asked the question, and understand why you'd hesitate to try to summarize the sentence you quoted above. But citing claims is rigorously and deliberately tight when it comes to claims from these kinds of sources. Professor marginalia (talk) 00:49, 31 December 2008 (UTC)
Sure. But in each of the cases I proposed above, I think the generalizations in question would remain very close to the findings. And we run into a major problem here as well - we very rapidly start trading verifiability by all editors (who can check, for instance, that the words "distal colon" do appear in the original) for comprehensibility to all readers (who are much helped by the loss of specificity involved in just saying digestive system.) Jargonny language is cautioned against, and with good reason. In practice we want to chart a middle ground. And I would suggest that the no specialist knowledge sourcing articles are taking their cue from NOR, since the "no specialist knowledge" phrase recurs verbatim - one imagines that a change here would trickle down. But as a reader, an article that refuses to substitute terms that are hyper-specific for the purposes of scientific precision with general terms for a less knowledgeable audience is unhelpful in the extreme. Phil Sandifer (talk) 01:42, 31 December 2008 (UTC)
No, readers are not served by a "translation" of primary scientific research specifically intended for easy digestibility for the general public. Especially when it comes to the sciences, that's a definitely not a DIY editorial job, and would fill the science pages with more junk information rather than good. Professor marginalia (talk) 01:58, 31 December 2008 (UTC)
Well, it's a balancing issue - if, for instance, it is accurate to replace "distal colon" with "digestive system," that is useful. We need to not overdo it, but our job is not to take scientific articles and re-arrange the words into a new order, then declare the information free. Phil Sandifer (talk) 02:09, 31 December 2008 (UTC)
That is possible. After all, we are, separate from this, having no end of disagreement about the description/interpretation distinction. :) Phil Sandifer (talk) 21:33, 30 December 2008 (UTC)
What, after all, counts as "specialized" knowledge. To me, "distal colon" is basic literacy that everyone should learn in junior high school, if only to be able to read advertisements. Translating it as "digestive system" is totally misleading--I am unable to imagine a context where the two would be acceptable equivalents--proper writing consists or wording the sentence to make that plain where in the d.s. the d.c. is, the first time the term is used, perhaps with a diagram. The art of encyclopedia writing is to be able to write an article that is comprehensible at several levels. What would be over-specialized is use of a term that would be understood only by a gastroenterologist without explanation. DGG (talk) 06:56, 31 December 2008 (UTC)
Well, first of all, I still don't know if I'm right about the DC being part of the digestive system (I deliberately haven't checked, going only on the fact that the colon, sans adjective, is). Though now that I think about it a second time, I'd guess excretory system is probably more correct, particularly given the other two organs mentioned in the passage. Now, if a compound aids in the function of those three things, it seems likely to be a not-inaccurate summary to simply say that it aids in the function of the excretory system. But this gets to the problem - if you asked me whether a summary of "compound aids with the digestive system" and the original said those three organs, I would plausibly fall for it. And remember - we're working with one sentence here. One big problem is that summary usually involves trying to combine multiple sentences into a shorter summary - which is going to necessitate further condensing. Which is going to require more small leaps and contractions that I wouldn't be able to verify. Phil Sandifer (talk) 14:08, 31 December 2008 (UTC)
By far, the best bet when a source uses a specialized technical term is for us to stick to the terminology that was used in the source... That way there is no possibility of introducing our own misunderstandings of what the term means (a form of inadvertant OR) into our article. Blueboar (talk) 15:36, 31 December 2008 (UTC)
This causes a host of problems - the inability to use synonyms forces us into line by line word re-orderings that amount to plagiarism or copyvio. It also means we fail at our task of, you know, explaining and teaching, and making information available. If our relation to scientific literature is bare paraphrases of copyright protected articles from proprietary databases, we are in some rather serious trouble. Phil Sandifer (talk) 16:07, 31 December 2008 (UTC)
Phil, judging by the recent argumentative tacks, you appear to me to have lost perspective. If you're trying to be perfect, please re-read WP:BE BOLD. On the other hand, if you're trying to figure out the precise boundary between WP:Summary style and original research, perhaps it's best to just play it safe and stick with sources that can clearly be shown to be summaries of an issue you're writing about. ... Kenosis (talk) 16:34, 31 December 2008 (UTC)
I don't even understand what you're trying to complain about here. Phil Sandifer (talk) 16:50, 31 December 2008 (UTC)
I'm not the one doing the complaining. You've tried numerous angles of questioning and argument, picking away at the edges of something that has already been gone over repeatedly on this talk page, and which is already in the archives. In other words, there's nothing new here. It's been at extreme length to the extent that it's not so much a substantive discussion that genuinely invites other participants conversant in WP policy, but rather has become something where most observers would say, simply, TLDR-- fundamentally a lengthy series of "what if's", with interspersed criticisms about the logic of the policy. Of course you're entitled to do so, as I'm entitled to urge bringing it to some reasonable sort of closure that respects the lengthy work others have put into arriving at the present expression of this policy. The policy is clear. WP doesn't do the same thing that academics do. In fact it's rather the opposite-- we don't do original research, but rather WP:Summary style (except in the case of "lists" and the like). ... Kenosis (talk) 17:10, 31 December 2008 (UTC)
I understand why you're posing the questions as you are, PS, but they're not hitting the mark and these are some reasons why. You're taking a single sentence plucked from an abstract of a lab study and trying to rewrite it so it isn't plagiarized or as jargony. I know you're doing this to illustrate the argument you're trying to make, but there are larger reasons why the examples don't work. What kinds of generalizations can be drawn from primary research in the sciences are decided in the field of science, not some publisher's editorial desk writing for a general audience. Authoritative secondary sources in the sciences are what WP depends on for these kinds of generalizations in the science and medicine articles, especially textbooks and science periodicals (which write secondary analysis about primary research as well as publishing primary research). Yes, sometimes the secondary sources will still sound jargony--but you'll find here at WP the cited claims made from them are jargony as well. WP cannot act the role of "translator" of highly technical, specialized information--we stick to the sources. If the sourced claim is highly technical or specialized, the claims attributed to it need to be as well. This isn't much of a hurdle at wp for several reasons. Few claims made in published lab studies will warrant an article here for the general public that haven't also been covered in secondary sources. If you take a single jargony sentence from an abstract of a lab experiment to try and cite some more generalized claim, that's almost always very bad editing - cherry picking fringe pushers do this kind of thing, not responsible editors . Professor marginalia (talk) 17:06, 31 December 2008 (UTC)
Well said, both Kenosis and PM... I would say your points are valid when dealing with the Humanities as well as the Sciences. Blueboar (talk) 17:20, 31 December 2008 (UTC)

random break - an aside

I'm sorry, are you saying the situation would get better if I took multiple sentences or entire articles? I should think it would get dramatically worse. The fact remains - I am a reasonable, educated person, and I am fundamentally unable to verify any summary of a technical scientific article that extends beyond simple grammatical rearrangements. No use of synonyms, no condensations, no distillations, nothing. I cannot verify anything other than simple grammar games. I mean, if you think I can, by all means, show me the sort of summary of a technical source that you think is both a quality summary and verifiable by me. Go find one. If this is how Wikipedia works, you should be more than capable of it. Because I hold that the level of summary I am capable of verifying is well below what we can use in a useful manner. So please - show me an article where technical sources are summarized to your liking. Phil Sandifer (talk) 17:45, 31 December 2008 (UTC)
Can you give us a specific example of a Wikipedia article that contains a summary of a technical article that you are unable to verify? Not a "what if" hypothetical... one that is currently is Wikipedia? Blueboar (talk) 17:52, 31 December 2008 (UTC)
From your earlier example, PS. epithelial sodium channel= "It has been suggested that it may be a ligand-gated ion channel." cited to this. The term "ion" is added, which a term learned in a basic high school chemistry class and it's clear from reading the link they're talking about sodium ions. Another: the claim: "ENaC consists of three different subunits: α, β, γ." from the cite: "ENaC is a heteromultimeric channel that is made of three homologous α, β, and γ subunits that share approximately 30% homology at the amino acid level". Professor marginalia (talk) 18:16, 31 December 2008 (UTC)
LOL. But PM, you're not just any old run-of-the-mill "reasonable, educated person without specialist knowledge". ;-) Wishing everyone a pleasant New Year. ... Kenosis (talk) 18:23, 31 December 2008 (UTC)
When it comes to Chemistry... I would say I qualify on the " without specialist knowledge" front (I didn't even take Chemistry in high school), and yet I had no trouble verifying those statements. Blueboar (talk) 18:36, 31 December 2008 (UTC)
Whereas I'm cautious - I do not feel that I can verify the validity of adding "ion" to that summary. In the article in question, incidentally, footnote two is one I outright cannot verify - it seems to me like the source in fact does not verify it, as it suggests that ASIC is a subcategory of ENAC, which our article suggests would be wrong. So I'm far from consoled on this point. Phil Sandifer (talk) 22:39, 31 December 2008 (UTC)
Chemistry for kids, "You are a sodium ion (Na+)".[4] The abstract wrote (Na+) instead of "sodium ion". This is not a substitution that requires specialist knowledge. Footnote two is to an article in Nature magazine that I can't get into today. I can't give an opinion because I can't read the article. Professor marginalia (talk) 23:34, 31 December 2008 (UTC)
That's not actually the problem I'm having. Even if we grant that ions are non-specialist (and I'm going to guess that most people would have trouble with what an ion is), the main reason I fail to be able to verify it is not that I don't know that NA+ is an ion - it is that NA+ is not paired with the word "channel" anywhere in the cited source. And so, while I will grant that NA+ is an ion, I am unable to confirm that it is correct to refer to this as an ion channel. And there's the problem. Ions have something to do with what's going on in that abstract, but it is not clear to me that what is going on is that there is something called an ion channel. Phil Sandifer (talk) 23:50, 31 December 2008 (UTC)
I think it's that you're out of your comfort zone rather than lack specialized knowledge. It's not ambiguous what they're talking about. The "pairing" is evident in the title, "Epithelial Sodium Channel: A Ligand-Gated Channel" Professor marginalia (talk) 00:38, 1 January 2009 (UTC)
This seems like a splitting of hairs - I am not comfortable verifying the summary. It is certainly not easily verifiable for me. Phil Sandifer (talk) 01:11, 1 January 2009 (UTC)

To clarify on my most recent comment here, as it was written on New Year's Eve and is un-fleshed out, yes, I am out of my comfort zone. It's possible for me to look at the description and see that it is plausible. But I do not think that I can get as far as verification. I can say that it is not obviously wrong. But no better. To verify it, I have to make the following moves.

  1. I have to associate Na+ with ion. Although this is elementary chemistry, it's something I bet more people wouldn't know to do than would. As it happens, I can do that step - but then, I'm married to a chemist, so I'd be sleeping on the floor if I couldn't.
  2. I have to grant that the move from a specific ion - Na+ - to the general statement "ion" is acceptable - that is, I have to grant that accuracy is not lost by the change in specificity.
  3. I have to grant that all uses of sodium in the abstract refer to the Na+ ion, and thus that the "Epithelial Sodium Channel" in the title refers to Na+ and not simply Na.
  4. I have to decide that the modifier "ion" can be validly placed after ligand-gated and before channel - something that neither it, sodium, nor Na+ does anywhere in the abstract.

I can do #1. But #2-#4 I can't. Now, I can't say that I don't follow those. They all seem plausible. But on the other hand, I can believe just as easily that you can't use Na+ and ion interchangably even though Na+ is an ion (just as you can't replace all instances of "steel" with "metal" in an article on bridge-building and retain accuracy), that Na+ is being used to be more specific than just sodium, and that there is such a thing as a sodium channel and a ligand-gated channel, but that a ligand-gated ion channel is just a made up string.

You've found a summary that is believable to a non-specialist. But it is not, for me, verifiable. If the current wording of NOR is to be read strictly, were the Horisberger/Charabi paper being used as a primary source, it could not be used to say that ENaC is theorized to be a ligand-gated ion chamber. Phil Sandifer (talk) 15:06, 1 January 2009 (UTC)

Several things need to be remembered here... 1) we are not talking about verification of the claims made in the source... we are simply talking about verifying that the source contains the information we are describing. 2) If, you have to look up some technical terms in order to complete this verification, that is fine... verification (even "easy" verification) is not the same as instant verification. In this case... since you can go to a standard chemistry text book (something that does not require a specialist knowledge to read) and see that Na+ is the the symbol for a sodium ion... or to look up any other basic bit of information you may not fully understand... then the article's description of what is contained in the source can still be "easily verified" by anyone without a specialist's knowledge of chemestry. The same goes for a description of what is contained in any other source. "Easily verifiable" does not mean instantaniously verifiable! Blueboar (talk) 15:53, 1 January 2009 (UTC)
I am aware - the issue is that the phrasing in question does not appear in the linked source, and so I cannot actually verify that it says precisely what the article claims it says. The move to a standard chemistry textbook is good enough to get me the Na+ is an ion claim - but as I said, that is, of the four steps, the least problematic, and the one I am at least willing to grant, albeit skeptically - I do not know how I would go about verifying that steps #2-4 are valid in a fashion that could fairly be described as easy. Phil Sandifer (talk) 16:07, 1 January 2009 (UTC)

Peer-reviewed reports of experiments: primary or secondary?

From the archives:

Primary sources present information or data, such as
  • archeological artifacts
  • photographs (but see below)
  • historical documents such as a diary, census, video or transcript of surveillance, a public hearing, trial, or interviews
  • tabulated results of surveys or questionnaires
  • written or recorded records of laboratory assays or observations
  • written or recorded records of field observations
  • artistic and fictional works such as poems, scripts, screenplays, novels, motion pictures, videos, and television programs.
Secondary sources present a generalization, analysis, synthesis, interpretation, or evaluation of information or data from other sources.

My questions are: (1) Is a peer-reviewed report of a laboratory experiment a primary or a secondary source? (2) What criteria should be used to examine borderline cases? 69.228.201.125 (talk) 08:13, 31 December 2008 (UTC)

Here are some reliable sources on the issue, wrt science/medicine:
Other fields may differ in their definition of PSTS. The first publication of original research (e.g. in a peer-reviewed journal) is considered a primary source (indeed, the lab books are irrelevant for Wikipedia, since they are unpublished). A review of the literature (also to be found in a peer-reviewed journal) would be considered a secondary source and is often a better source for our purposes. Some sources may contain primary and secondary material, or can be considered primary/secondary depending on what facts are being drawn from the source. Borderline cases are probably best discussed in context with the material (Wikipedia text) they are supporting. Colin°Talk 08:49, 31 December 2008 (UTC)
Interesting! Thank you! Would an academic journal article which reports on an experiment and includes an analysis, interpretation, and an evaluation of the data collected from it qualify as both primary and secondary? 69.228.201.125 (talk) 09:52, 31 December 2008 (UTC)
Thinking "X interprets Y, so X is secondary, Y is primary" is too simplistic. Data goes through many stages of interpretation/evaluation before it reaches a log book and again before it reaches publication. When a source is based on previously published sources, it becomes secondary. In a research paper, the focus is on the original research and this is what most folk will cite the paper for. However, such a paper often contains a short discussion of prior work/research in the field: this is secondary material. However, unlike a dedicated literature review paper, such content is seldom cited on WP as it is not the focus of the paper. The analysis, opinions and conclusion in a research paper need to be read in context. For example, basic lab research in the health sciences should not be given clinical relevance by WP editors, no matter how important the researcher thinks his discovery is. Colin°Talk 11:06, 31 December 2008 (UTC)
There's no hard-and-fast dividing line. If in doubt about what's primary and what's secondary, try to find sources that are clearly secondary or tertiary so as to facilitate WP:Summary style. Only if there are no secondary or tertiary sources on a particular issue should primary sources even be considered. If there are no sources that clearly are not primary sources, ordinarily that should be a hint to double-check the issue's standing per WP:Notability. ... Kenosis (talk) 16:12, 31 December 2008 (UTC)
I really think it would help if we included a clear statement to the effect that the definitions of Primary/Secondary are different for different accademic fields, and that a single source might contain both primary and secondary material within it.
I also think we need to remember that the intent of the PSTS section isn't really to clearly "define" what is primary and what is secondary... the intent is to warn editors that: because it is easy to misuse primary sources (in a way that introduces our own Orignial Research into an article), we need to be extra-cautions when we do so. Blueboar (talk) 16:17, 31 December 2008 (UTC)
The statement in the policy about use of primary sources is already clear. Remember, these are editorial policies that allow contributors and editors a great deal of good-faith discretion across the wiki. Si, if in doubt, use the extra measure of care and follow the primary-source rule w.r.t. all kinds of sources. If you've got a piece of arcane material in front of you that you can look at and say "I think that with respect to "issue X", I'm looking at its origins, not just some watered down summary", then most likely you're doing what is understood in WP to be original research. In such a case, with respect to WP at least, you should to be looking for summaries of what this arcane material may be talking about. If you can't find any such summaries (secondary or tertiary material), then ask yourself the question whether it meets the WP inclusion criteria. And, if you're that determined to use it, use it per WP:BE BOLD. Nobody gets arrested around here (to date at least) for violating WP:PSTS, or WP:SYN, or WP:NOR, or even for obvious violations of WP:NPOV. If there's disagreement with another user, take it to talk and discuss what you're writing about in the context of the availiable literature on the particular issue -- not here necessarily, preferably on the article talk page. ... Kenosis (talk) 16:50, 31 December 2008 (UTC)
I was thinking more along the lines of including a clear statement about different discipines using different definitons. Blueboar (talk) 17:11, 31 December 2008 (UTC)
You mean in the body text of the policy rather than in footnotes? I personally think a two-clause sentence or two short sentences about this would be appropriate. For example: "The dividing lines between primary, secondary and tertiary sources are not hard and fast, and can vary somewhat according to different areas of study.[provide two or three footnoted examples, such as historiography, scientific research, or whatever, plus one or two standard lists emanating out of multi-disciplinary library science]" ... Kenosis (talk) 17:31, 31 December 2008 (UTC)
Yes, something like that. I think it would be helpful. We often get arguments along the lines of "but that isn't how <Insert accademic discipline here> defines primary/secondary source". We should acknowledge this (if only to end the arguments). The point being that it does not matter which difinition we are using ... the cautions and caviats we talk about in the policy are the same for all of them. Blueboar (talk) 17:59, 31 December 2008 (UTC)
Well given that secondary is best (in general), it does matter that WP editors agree on which sources are secondary rather than primary or tertiary. I think WP should pick a definition that serves our purposes and use that. We can link (via footnotes) to sites (like those above) that help editors with PSTS where those sites have a definition consistent with WP. By all means acknowledge that some folk use different definitions, but it helps if we all sing from the same hymn sheet. Wrt the above two-sentence clause, it misses out that sources can be a mixture of primary and secondary material, and what a source is used for can also influence whether it is primary or secondary material. Colin°Talk 19:48, 31 December 2008 (UTC)
It's not necessary to achieve universal, precise agreement in advance regarding the exact dividing lines between primary, secondary and tertiary. There are not precise boundaries between them and every field has some exceptions to the examples commonly given. The language of the policy allows more than adequate discretion and very wide latitude for editors to just go about their business of being unpaid contributors or editors, so long as they're not doing original research in WP. Where there's disagreement about what's what in terms of WP:PSTS, or about where article writing ends and where original research begins, editors can readily discuss it at individual article talk pages.
..... About the observation that a single source can be regarded as primary w.r.t. one issue and secondary w.r.t. another, it's a good point and I think should probably also be concisely noted in conjunction with a statement of the kind mentioned just above ... Kenosis (talk) 21:50, 31 December 2008 (UTC)

I would suggest that we ought not redefine terms that have accepted meanings with our own meanings. If we are intent on doing so however, I think there is something to the idea that something that has been peer-reviewed and edited as necessary is fundamentally not in the same category as an eyewitness report or a novel. I think that peer review may constitute a significant enough step to mitigate against primary sourcing. Phil Sandifer (talk) 15:02, 2 January 2009 (UTC)

"I would suggest that we ought not redefine terms that have accepted meanings with our own meanings." I completely agree... which is why the policy is written the way it is... we may describe what a source says, sticking as closely to the language used in the source as possible... and should not attempt to interpret what the source means. We stick to the source! Blueboar (talk) 17:01, 2 January 2009 (UTC)

Chemistry verification

I just pulled up ten random articles found by wandering about the chemistry categories and clicking on things I didn't know what were. My goal was to see how true it is that our use of technical sources is verifiable by non-specialists. These were random picks, though I threw back articles that had no citations to technical documents that I could access via the web. I did not throw back articles where I could verify all the citations I tried to verify.

  • Grotthuss mechanism - footnote 5 is in no way verifiable from the abstract linked to. Going to the full article proves unhelpful.
  • Aurophilicity - footnote 3 is plausible, but I am unable to confirm some key portions, most notably that it is acceptable to equate "Pd(II)→Hg(II)" or "Hg−Pd" (I can't tell which one is applicable) with "Hg(II)···Pd(II)". I am assuming that the footnote is only meant to verify the Pd/Hg portion of the sentence, as the others do not even appear at all.
  • Debye length - I cannot find the equation given in any of the three sources.
  • Eutectic point - I was fine with this one.
  • Jahn–Teller effect - Although aspects of references 3 and 4 can be verified, the exact phrasing here is problematic.
  • Constrained geometry complex - Well, the good news is that the reference is verifiable. The bad news is because it's a word for word restatement.
  • Paterno-Buchi reaction - Reference #2 is completely beyond me.
  • Resonance fluorescence - The reference here appears to actively be wrong to me.
  • Slater-Condon rules - Nothing in reference 4 suggests that it follows from Condon's work.
  • Coelenterazine - Ref #1 is problematic for me - the ref suggests to me that extracts were added to coelenterazine, suggesting in turn that it is not naturally occurring. This is a borderline case.

It should be noted, far harder than finding technical references was finding references at all - our scientific articles remain largely unreferenced, and current policies suggest that the appropriate thing to do is to hack them to pieces. But beyond that, when references to technical sources exist, those references are often extremely opaque to non-specialist readers.

I am hard-pressed, therefore, to believe that this standard of easy verification by non-specialist readers is operative or has ever been operative when it comes to sources that are aimed entirely at specialist audiences. It continues to be my sense that the most appropriate fix to this problem is to add a note that descriptive claims about sources that clearly meet our standards of reliability and are aimed at specialist audiences are allowed. Phil Sandifer (talk) 16:04, 1 January 2009 (UTC)

Lots of stubs and unclassified content, some are simply not articles at all. I think you would be better served by presenting examples that have been reviewed and are classed as GA or higher. Semitransgenic (talk) 17:12, 1 January 2009 (UTC)
Well, though you have to ask what the status of this policy is if such a massive swath of articles and coverage violates it. Our policy is not generally taken to be spitting into the wind. That said - the citation at the very end of the "Induction" section of the article Long-term potentiation is unclear to me - it does not appear to match what the section of the source talking about MAPK is talking about. Footnote 37 of Persistent carbene is also very much unclear to me. Those are what I find with a few minutes effort. Again, though - if widespread violation of this policy is accepted (as I do not think you would get very far deleting any of the ten articles listed above, nor most of the ones I found and skipped over because they were wholly unsourced. Though feel free to try) then I think we need to reconsider how we phrase things. I suspect that a note regarding reliable sources aimed at technical audiences and perhaps an acknowledgement that easy verification is a goal, not a requirement would best reflect actual accepted practice. Phil Sandifer (talk) 18:15, 1 January 2009 (UTC)
Phil, you are missing the existence of a fundamental principal: concensus. I see the point you are making, I tried to make a similar point regarding the OR threshold on this very page, but I now accept that, in a nutshell, if an article succeeds in passing the criteria for speedy deletion, doesn't hit PROD or AfD, it's most likely going to stay around for a while; eventually someone might dispute it's content, demarkate it for improvement, of one description or another, or actually attempt to improve it themselves. Changing the specialist knowledege provision will not alter any of this. If the various articles in question really bother you so much I'm sure it would be appreciated if you actually set about tackling the existing issues on a case by case basis rather than wasting time here trying to get a result it does not look like you are going to get : ( Semitransgenic (talk) 19:39, 1 January 2009 (UTC)
Well, I think one thing we need to be cautious about is the difference between consensus on this talk page and project-wide consensus. Policy pages walk a delicate line between setting practice and describing practice. There are, what, half a dozen of us discussing this, and 2.5 million articles out there that this ostensibly governs. And right now, on this point, there's a real disjunct between theory and practice. And this does become a problem, because when the theory prescribes something other than practice, it leads to cases where the theory is abused. We need to be careful with inaccurate policy pages, as they tend to get used by rules lawyers and by mindless process wonks in destructive ways.
As it happens, the science articles don't bother me. I think they're undersourced, but I don't see them as urgent problems requiring fixing. I don't know for certain that they all summarize their sources correctly, but I don't know that they don't, and it seems foolish for me to start raising objections in article areas I don't know anything about. In fact, that is where my problem arises - I think consensus is, generally, that articles such as the 10 listed are OK, and that the problems in verification I found quickly on the two GAs are also OK. That makes me think that the proposed idea that all citations, whether to primary or secondary sources, must be easily verifiable by non-specialists is a non-starter - because that is clearly not current practice on any level. Though at this point we may also be suffering from discussion drift. I'll try to refocus it later today and see if it's possible to make it move forward. Phil Sandifer (talk) 20:35, 1 January 2009 (UTC)
Notes on examples:
  • Grotthuss mechanism - An author of the paper cited added it to the article himself.......this is, obviously, a poor example of "standard editing" at wikipedia since the editor edited so little here except to add his own published work in cites on wikipedia. This is a unique case raising different issues.
  • "Aurophilicity - I am assuming that the footnote is only meant to verify the Pd/Hg portion of the sentence, as the others do not even appear at all." I can't access this article either but it's my guess we have a synth problem there. Is the term aurophilicity used in the article? Is there any mention of Au(I) at all? It smells like a synth, but I can't say for sure because I haven't read it.
  • Debye length - If this is the case, it's not sourced then. Simple.
  • Eutectic point - Apparently ok
  • Jahn–Teller effect - Isn't #3 published in German? And does #4 directly claim to contradict the findings in #3? I didn't find evidence that #3 was cited in it. What is the wording in #4 that addresses the #3 paper? If it isn't addressed directly, it's original research.
  • Constrained geometry complex - If it's plagiarized, that's another issue
  • Paterno-Buchi reaction - #2 is not the best reference to use there, I'd agree. If the phenomenon was truly identified at least 100 years ago there should be more suitable sources.
  • Resonance fluorescence - "The reference here appears to actively be wrong to me." I think you're right [5] That stub certainly needs work. :)
  • Slater-Condon rules - If nothing follows, it's synth then. When primary research is cited as an "example" of some principal but doesn't address the principal, you need a third party source to make the association. Synthed claims are clearly not allowed at WP.
  • Coelenterazine - I think you misread it, PS. The extract process was done to test for coelenterazine in the phylum chaetognatha, and the author(s) explicitly confirmed coelenterazine was the source of its luminescence. I don't think the "copepod" was explicit in the ref-strictly speaking it would need one too, which probably isn't difficult to do.
I'm not completely convinced yet, since some of these comments are based on my hunch (as I noted) of original research problems with the cites. But at this point, I think only one of the list above might illustrate the problem with the "specialized knowledge" clause-that one being the Paterno-Buchi reaction. And as I said, I think a better source could and should be used on that one. If we're indeed finding so many cases of synth, does it make any sense whatsoever to relax this clause? We need to be more careful, not less. Professor marginalia (talk) 22:09, 1 January 2009 (UTC)
Well, and here's the question that makes me cautious here. For instance, in the Slater-Condon rules, I do not know the status of Condon's expansion upon Slater, and if the third source in question actually does expand on Condon. It may well be that any response to one is implicitly a response to the other. And thus that this is not synth. If everybody who is actually reading journal articles about the Slater-Condon rules would know that responding to Slater's rules means that you're responding to Condon as well, then no source is going to make that explicit. Which is the problem we're running into - these sources are sources that are not meant to be read by non-specialists, and have taken no care to be readable by them. I'm very much skeptical of the benefit of trying to treat them as anything other than what they are. Phil Sandifer (talk) 22:17, 1 January 2009 (UTC)
So the question you have is simply whether Lowdin can be said to have generalize from Condon's expansion also? This secondary source would better sort the confusion then. This is an example of the inferiority of the primary source-the cite is to the expansion paper itself, which is good information, but at WP it's not the source for citing claims about any association to Condon's contribution. Again, this is why secondary sources are invaluable. What these examples are illustrating is that this "I can't judge because I'm not an expert" deference to just about any claim cited to a highly technical document will let pass all kinds of inadequately or even incorrectly sourced claims! If it doesn't look right, it's very likely not right. Professor marginalia (talk) 23:50, 1 January 2009 (UTC)
See, whereas I'd suggest that part of the issue here is that I shouldn't be editing Slater-Condon rules to begin with. Which is, I think, a valid question to raise - if key sources in an area require specialist knowledge, should non-specialists who don't understand the sources be editing those articles to begin with? And I suspect the answer here is no. Which isn't to say we need to create any sort of policy of credentialism - far from it. But on the other hand, I don't know that we have much to gain for explicitly providing for editors who it is probably undesirable to have editing the article in the first place. In a discussion about how Slater-Condon rules should look, I should not have an equal vote to someone who has actually, you know, heard of them. :) Phil Sandifer (talk) 14:55, 2 January 2009 (UTC)
"if key sources in an area require specialist knowledge, should non-specialists who don't understand the sources be editing those articles to begin with? And I suspect the answer here is no." Actually, you could not be more wrong on this... the answer here is unabashedly "YES!" Wikipedia was founded on the concept that "anyone can edit". While we welcome the contributions of specialists, they have no special claim to editing superiority. Nor are they in anyway exept from the policies and guidelines that govern the project (nor should they be).
I think this may be part of the difficulty we are having here... you seem to want Wikipedia to be written by specialists for specialists. However, that is not what Wikipedia is. Blueboar (talk) 16:53, 2 January 2009 (UTC)
Indeed, anyone can edit. However, there's an is/ought distinction at play here. Wikipedia does not check credentials or require credentials, nor should it, nor do I see it as likely that it ever will. That, however, is a different statement from saying that anyone should edit. Is anyone going to check my credentials at the door before I can edit Slater-Condon rules? No. Is it a good idea for me, having no idea what the Slater-Condon rules are, and having none of the fundamental knowledge needed to understand them, to start editing the article? No. The fact that we do not check credentials does not mean that we encourage people to edit topics they don't actually know anything about. In fact, we ought to discourage it. Just not via credentialism. Phil Sandifer (talk) 17:12, 2 January 2009 (UTC)

Attempting to summarize discussion

The discussion of the "no specialist knowledge" clause has been wandering a lot, and I'm finding that I'm having trouble keeping track, in a given thread, of exactly what issues are and are not at stake. So I figured it might be time to back up and try to sketch out what's actually under discussion and see if there are ways to work practically on solving the raised problems instead of getting side-tracked.

The underlying issue is this: because NOR treats primary and secondary sources differently with regards to specialist knowledge, it makes NPOV difficult on biographical articles of people who work on specialist topics, as criticism of the person's work makes it in under easier standards than the person's explanations and defenses of their work.

I may be wrong, but I think, at this point, pretty much everybody in the discussion at least sees that some sort of problem exists in this area - that in the situation where Person A is criticized by Person B and responds, we do not want Person B's criticism to be summarizable freely and not Person A's response.

Here are the solutions I've seen proposed, with some commentary on what has been said about them.

  1. Remove the "no specialist knowledge" clause.
    At this point, this seems off the table. Although it would solve the problem, it is drastic, and seems likely to have unintended side effects.
  2. Extend the "no specialist knowledge" requirement to all sources.
    I have raised several objections to this - first of all, it seems to me to result in the neutering of specialist POVs, and thus to create a larger NPOV problem than the one we're trying to solve. Second of all, it does not seem to me to describe current practice at all, and I do not think its enforcement on most articles would find consensus.
  3. Carve out an exemption for the summary of reliable specialist sources.
    This has not been discussed in great detail, but has at least not been strongly opposed. The idea would be to use structures such as peer review to help hedge against the problem of highly specialized sources, recognizing that they do have a role to play, but that we need to make sure their use is limited to descriptive claims about extremely reliable ones.
  4. ???
    That is to say, more ideas are much desired.

Does this seem like a fair account of where we are? If so, I'd be interested in having more discussion about #3 - it seems to me like a promising approach - one that deals with the reality that, in a comprehensive encyclopedia, specialist topics are going to need to be grappled with. The past few days of discussion about easily verifiable summaries have only made me more convinced that we have serious issues with easy verification of claims about specialist sources - by and large our use of specialist sources cannot be verified by non-specialists, and I think even in what are apparently simple summaries (like the ligand-gated ion channels example several sections above) there are enough problems to prevent "easy" verification by someone who does not have a handle on the material.

But any such exception is going to have to be very carefully delineated - both in terms of the sorts of claims that can be made (which seem to me to be in the category of what we're currently calling descriptive claims) and the sorts of sources that they can be made about (we want to limit this to rock solid reliable sources.)

Thoughts? Or other routes to pursue? Does this seem like a fair summary, if nothing else? Phil Sandifer (talk) 20:52, 1 January 2009 (UTC)

I think this is a fair summary of the issue and the arguments. My personal feeling is that you are making far more of the issue than it deserves (in other words, I don't think this is really a policy level problem).... but if we must deal with it, we should be looking for the solution by exploring #2 or #4 of your options. Blueboar (talk) 22:44, 1 January 2009 (UTC)
I think we do need to deal with it. I don't think #2 is feasible, as I said. What is your problem with #3? Phil Sandifer (talk) 23:08, 1 January 2009 (UTC)\
I don't think an exemption is justified. All sources should be treated the same in this reguards. Blueboar (talk) 04:14, 2 January 2009 (UTC)
Well, though since primaryness and secondaryness are not inherent properties of sources, whereas we generally do treat reliability as being inherent, I think I'd argue that treating reliable sources as reliable is more consistent, not less. Phil Sandifer (talk) 14:57, 2 January 2009 (UTC)
I disagree with the premise that primary and secondary sources are treated differently. I think it's a misinterpretation that may be caused (though I'm not convinced) by the extra emphasis applied to primary sources. Professor marginalia (talk) 23:28, 1 January 2009 (UTC)
Not convinced by any of the examples. See no pressing need to alter this policy. Semitransgenic (talk) 12:11, 2 January 2009 (UTC)
Well, to be fair, I think that misinterpretation is pretty fair, given that nothing resembling the no specialist knowledge claim exists for anything other than primary sources. Phil Sandifer (talk) 14:57, 2 January 2009 (UTC)

Another possible idea occurs to me, and I figure this is as good a section as any to put it in:

What if we loosen the standard from verifiability to falsifiability here? That is, if the issue is that we are choosing between two perspectives on a source: "Source X says Y" and "Source X says Z," that is a fundamentally different issue than the two perspectives "Source X says Y" and "I do not understand Source X." In the former case, we have an OR issue and need to move to secondary sources or find a statement other than Y and Z that all parties can agree on. In the latter case, there is no OR issue. Phil Sandifer (talk) 15:03, 3 January 2009 (UTC)

Verifiable... or "easily" Verifiable?

As an offshoot from all the conversations we have been having... while I strongly agree that any statement included in Wikipedia should be verifiable by any reasonable, educated person without specialist knowledge, I am less happy about saying that a statement has to be "easily verifiable". Verification is not always "easy" (for example, if a primary source is a somewhat obscure printed book, it may require going to several libraries before a descriptive statement about it can be verified). I think we either need to cut the word "easily", or we need to better explain what we intend here, what we mean by "easily verifiable". Blueboar (talk) 14:42, 2 January 2009 (UTC)

I would support the dropping of "easily." Phil Sandifer (talk) 14:49, 2 January 2009 (UTC)
ditto. How did it ever get in? DGG (talk) 17:21, 2 January 2009 (UTC)
OK, I have removed it... we will see if it sticks. Blueboar (talk) 17:57, 2 January 2009 (UTC)
Given this change, I think the "under discussion" tag should be removed. The "specialist" language has been in the policy for many 3 years, and has been reiterated in several other guidelines, essays, and arbcom decisions. Professor marginalia (talk) 18:10, 2 January 2009 (UTC)
I continue to object to removing it while there's an active discussion going on. Phil Sandifer (talk) 21:56, 2 January 2009 (UTC)
I believe you're the only one to have weighed in yet that wants to change it. The tag is unlikely to generate much input. How many experienced editors are likely to read it there? If you want to generate input, I suggest you put together an RFC. Professor marginalia (talk) 22:09, 2 January 2009 (UTC)
This discussion has suffered from having only a very few participants and a very convoluted chain of events and complicated examples. You four or five are not the only people who care about it. I am tending to agree with Phil that there are remaining problems with the policy, and I think that I speak for many other people in the community.
Policy being set by a small group arguing with themselves in the corner, especially policy this important, is poor practice.
So, please leave the tag up, while some of the rest of us come in and come up to speed... Georgewilliamherbert (talk) 23:00, 2 January 2009 (UTC)
I also support dropping "easily"; it's always been the case that we prefer sources to be easily available, but do not require it. Some facts simply aren't available from other sources. Dcoetzee 23:04, 2 January 2009 (UTC)
This could impact on a discussion I've been having with someone using a book by (from memory) the Inner Mongoloia Pubishing company which is not available so far as I can see in the West. It's technially verifiable, but not practically verifiable. Maybe we could have a couple of examples of what we mean? dougweller (talk) 07:07, 3 January 2009 (UTC)