Wikipedia talk:Scientific citation guidelines/Archive 2

Archive 1

Archive 2

Annotations section

As noted above, the 'Annotations' section may need some work. It's confusing on the first read though I've found after going over it a few times it does start to make sense. I think there are more serious issues though. First, an annotation is meant to be supplemental material and is therefore very different from a citation. So it seems doubtful that a guideline on annotations belongs here at all. Part of the section seems to be trying to encourage adding explanatory remarks to long lists of of references or further reading lists so readers have an idea of what to expect without looking up each one. An example of this is section of Simplex algorithm#Further reading, a typical entry is

• M. Padberg, Linear Optimization and Extensions, Second Edition, Springer-Verlag, 1999. (carefully written account of primal and dual simplex algorithms and projective algorithms, with an introduction to integer linear programming --- featuring the traveling salesman problem for Odysseus.)

The annotation is in parentheses here. These may be helpful but they could lead to verifiability and POV issues in themselves, e.g.

• J. Smith, Linear Programming, Acme Press, 2010. (the very best linear programming book ever written!)

The example given does not seem me what we should be encouraging as a best practice. It appears in the text to be a citation but when you read the footnote it's just a list of books and doesn't support the statement at all. If this was an article I was working on I'd move the footnote to a 'Further reading' section since it does not belong in the references section.

The last sentence talks about long annotations and I agree that if they are used at all then they should be kept short or moved into the main text. I'm wondering if we should be encouraging their use though. As example something like

The sun is very hot.^[1]

1. ^ 5,778 K

should probably be rewritten

The temperature of the sun is 5,778 K.

In any case this does not seem to be a scientific issue so it's unclear why it would be here rather than another guideline.--RDBury (talk) 01:47, 22 November 2010 (UTC)

Your example illustrates nicely the importance of sourcing scientific "facts". What do you mean when you say "the temperature of the sun is 5,778 K"? The spurious accuracy is already suspect. Do you mean the "surface temperature"? If so, what do you mean by the "surface"? The corona is considerably hotter. And closer to the center the temperature of the sun reaches millions of degrees K. So how can Wikipedia be so confident that the temperature of the sun is 5,778K? Geometry guy 01:57, 22 November 2010 (UTC)

One of the difficulties with this guideline is that it purports to be a source citation guideline but in fact discusses things that are nothing to do with sources in the way Wikipedia refers to them (reliable, typically secondary sources that directly and specifically support the article text), but have got mixed in because they often appear in footnotes. As noted above, the example in annotations section lists lots of textbooks. That would be unacceptable to anyone reviewing the sources for an article. They'd say "give me one textbook and can I have a chapter and page number". I can see merit in supplying an alternative source if the main one is difficult to get hold of or if the alternative is easier for the general reader to read. But the footnote mustn't turn into some sort of bibliography and comments about the quality of the textbooks (beyond perhaps saying if one is advanced, beginner, detailed, or general) have no place here: were are not Amazon! And, I'll repeat that I don't see anything "science" about footnotes or source citations that say a little bit more than just the minimum. Colin°^Talk 18:32, 22 November 2010 (UTC)

That all sounds a lot more like personal opinion and a lot less like something that we should be telling all Wikipedia editors to follow, to me. Only one text is allowed? Footnotes can only be sources? We are not allowed to say anything about the sources? Why ever not? —David Eppstein (talk) 18:42, 22 November 2010 (UTC)

I don't think that's what Colin is saying. This is supposed to a guideline on citations, guidelines on other things belong elsewhere. Anyway, removing that part won't imply you can't do what it suggests, it just won't encourage it anymore. After some thought it seems to me that if the section belongs anywhere then it would go in Wikipedia:Manual of Style (layout) since that's where the Further reading section is described. I'm going to go ahead and propose that it (or some trimmed down version) be added there. It doesn't seem to be safe to delete it from here yet.--RDBury (talk) 18:57, 22 November 2010 (UTC)

I think it is quite useful, when giving citations to textbooks, to give several books rather than one, if there are several common options> I also think that short annotation, like in footnote 17 of the current revision [1], are helpful. Of course they could be overdone, and I personally dislike too-subjective annotations. But matter-of-fact descriptions are fine by me.

The goal of references in an encyclopedia is not only to verify facts. Like a review paper, an encyclopedia article is intended to be a starting point for a reader, not an ending point. So our references are also intended to make it easy for readers to go further into the literature. Pointing out which references are textbooks, and which are suitable for undergraduates or for a general audience, is an important part of our role. For topics that are well covered by textbooks, we do our readers a disservice if we fail to point this out. — Carl (CBM · talk) 19:05, 22 November 2010 (UTC)

This concept, of using references to both meet WP:V/WP:NOR/etc and to supply an annotated bibliography for the reader's own research, is AFAIK, not general practice on Wikipedia, not followed by our best articles, and not documented elsewhere on policy or guideline pages. It also, if you allow me to repeat myself, is not peculiar to science articles. Other disciplines have textbooks too :-). Colin°^Talk 19:12, 22 November 2010 (UTC)

The requirements to provide both references for verification and references for further research is mandated by our goal of writing an encyclopedia.

If other disciplines for some reason place less value on the usefulness of their articles as entry points into the literature, that's fine. This page reflects the priorities of editors from certain WikiProjects about what we expect in our best articles. There is no reason that we cannot impose stricter requirements on ourselves than are required by WP:V. That is exactly what this page does. This page extends WP:V in various ways, but it does not in any way lessen the requirements of the general policies. — Carl (CBM · talk) 19:19, 22 November 2010 (UTC)

Leaving completely aside any merits of Carl's opinion, I do not believe that his assertion that policies require articles to contain references for the reader's further research is supported by either the community in general, or by any written policy page. WhatamIdoing (talk) 22:09, 24 November 2010 (UTC)

It's interesting to compare what's happening here with Citizendium which says that their references are for the reader's benefit, not to resolve disputes between editors. 89.241.232.14 (talk) 00:33, 23 November 2010 (UTC)

That's interesting - do you have a link to it? — Carl (CBM · talk) 00:41, 23 November 2010 (UTC)

The link is http://en.citizendium.org/wiki/CZ:We_aren%27t_Wikipedia section "How do we differ?", item 6 89.241.232.14 (talk) 00:59, 23 November 2010 (UTC)

There should be in-line citations only, unless those are not available. Bibliographies lead to long lists of arbitrary titles, the relevance of which for the article is not clear or absent. The real information gets lost into irrelevant information.--Wickey-nl (talk) 16:06, 24 November 2010 (UTC)

Even when the entire article is sourced to one or two single-page citations? This does happen; it probably happens more frequently in the sciences than elsewhere. To require in-line citations when they would all read "Smith and Jones, p. 645; for a category theoty approach, see Miller, p.987" is silly; citations are made to serve articles, articles do not exist for the sake of citations. Septentrionalis PMAnderson 05:58, 26 November 2010 (UTC)

This is not a Wikipedia guideline; it is just a Wikiproject page

This content guideline on citing sources has a vague scope, misleading title, extends into issues not pertaining specifically to science articles, deviates from policy, doesn't document best practice, and merely reflects the opinions of a few wikiproject members. It should be removed from the Category:Wikipedia content guidelines and placed into project space. Colin°^Talk 22:24, 22 November 2010 (UTC)

Specifically

1. It claims to be a Wikipedia content guideline on citing sources for scientific articles. But the comments on attribution and annotations (i.e., an opinionated bibliography) fall outside of this scope. The "Examples, derivations and restatements" is a guideline on WP:NOR rather than WP:V. It would probably be helpful if the scope was increased to include issues beyond just source citations.

2. Its title (scientific) is misleading. I'm getting the impression editors here write articles on mathematics and theoretical physics rather than experimental science. I appreciate there is a sub-heading saying "Mathematics, Physics, Molecular and cellular biology and Chemistry." but the title is still problematic. The discussion on the need to cite original sources highlights the difference between the type of "science" being written about here and the sort of science that most other disciplines follow. For example, the comment "there is no other source that can describe an original idea as faithful as the original publication." and "Well, ideas usually don't get worse over time.".

3. It covers ground that isn't specific to "science" articles. For example, whether to use inline citations and what the references section is for (sources, or bibliography). This is also relevant to the (now deleted) section on summary style.

4. It deviates from policy. For example, requiring the primary sources be cited "even if they are not used as sources in writing the article", and the emphasis on using original sources unless they have been shown to be erroneous.

5. It is the personal opinion of some editors at a few wikiprojects, rather than being a content guideline accepted by Wikipedia. (see comments like "This page reflects the priorities of editors from certain WikiProjects about what we expect in our best articles.") Indeed, Joke137 (talk · contribs) had the right idea when he created this page and flagged it with "Opinions are the author's and are not necessarily shared by WikiProject Physics, but they should be." The page has changed remarkably little since.

6. It does not document best practice. Our FA articles on "Mathematics, Physics, Molecular and cellular biology and Chemistry." do not follow the guidelines in this article. Specifically, the requirement to provide attribution for original thought or discoveries through citing the original works as sources, the mixed use of inline citations to support WP:V and to provide a bibliography.

I believe this page should be removed from Category:Wikipedia content guidelines and placed into project space. I think editors here actually want to document more than just their opinions on citations but also want to be left to agree among themselves rather than accept the limitations and outside influence that a formal Wikipedia guideline would impose. Colin°^Talk 22:16, 22 November 2010 (UTC)

For anyone coming here because of the RFC: please see the extensive discussion in the previous five sections of this talk page. —David Eppstein (talk) 22:27, 22 November 2010 (UTC)

Again, it's impossible to respond to 6 individual things in one section. There is clear precedent for subject-specific guidelines (this and WP:MEDRS, the guidelines on Asian naming conventions, etc.). It's true that the name isn't great but there's none better ("Guidelines for several science-related wikiprojects"?). — Carl (CBM · talk) 22:42, 22 November 2010 (UTC)

Well there wouldn't be six individual things if there wasn't so much wrong with it :-) Feel free to create sub-sections to respond to particular issues. Yes we have subject-specific guidelines (mainly MOS but some other areas). But these other guidelines document best practice which is clearly evidenced by the FAs that comply with them. And they don't deviate from policy or from other better established guidelines. They explain how to apply policy to that subject area. They don't create a set of rules that duplicate or deviate from the rest of WP. Colin°^Talk 22:56, 22 November 2010 (UTC)

Nothing in this page deviates from any policy. The only example you have claimed is that this page asks for citations to the source of a theory or experiment. But these sources are used in the process of writing the article: they are used as sources for the information about the original publication of the work. This is a permitted use of primary sources per WP:NOR, because the sources are used in a purely descriptive way, namely to describe the publication details. — Carl (CBM · talk) 23:08, 22 November 2010 (UTC)

This guideline specifically requests sources that have not been read be cited. Which is against policy and WP:SAYWHEREYOUGOTIT. WP:WEIGHT requires secondary sources. Otherwise, one might fill our articles with the all the random ideas, studies and experiments that get published over the years. Primary sources can only be used as an adjunct. Articles are based on secondary sources. That's what WP:V says. Colin°^Talk 23:15, 22 November 2010 (UTC)

Nowhere does this page explicitly say to use sources you haven't read (I looked). The expectation in all writing is that whenever you cite a source you'll look it up. So when the guideline says "To this end, editors of these articles should consider citing the original papers, even if they are not used as sources in writing the article." it means "you should go back and look them up if you left them out when you wrote the article originally". I will clarify that text, since you misread it.

Also, nothing on the page says to base the article on the original publications. Only to include them as supplementary references. I have added a quote in a higher section from the AMS ethical guidelines that motivates this further. — Carl (CBM · talk) 23:25, 22 November 2010 (UTC)

This RfC seems like sour grapes for failing to impose one's opinions in a few very narrow respects that are mostly MOS-like than anything else (although MOS-type stuff creating the biggest ruckus in Wikipedia doesn't surprise me at all.) So, I vote keep this as guideline because it's generally useful. Tijfo098 (talk) 06:56, 26 November 2010 (UTC)

I created this RfC because this guideline's text was in conflict with policy. Whenever there is a disagreement, it is always useful to bring in other editors. This wasn't "sour grapes". The guideline had serious problems, not some MOS triviality. I'm disappointed that editors from other disciplines didn't join the discussion and the RfC failed in that regard as I reckon most people here had this page on their watchlist already. The worst aspects of this guideline have been fixed so I've removed the "disputed" tag. I've also remove the RfC notice--If anyone disagrees with that, they can restore it but I'm done with it. Colin°^Talk 19:52, 26 November 2010 (UTC)

Examples from FAs

Regarding point 6 above, it is not difficult to find examples of featured articles that cite the sources of original publication of their topics. For example, General relativity, Aldol reaction, Quark, Nicotinamide_adenine_dinucleotide, Oxidative_phosphorylation, Euclidean algorithm. Often these are in the "History" section of the FA article, or in the lede. Whenever you're writing a carefully-sourced document that includes an historical survey, which could be an encyclopedia article, a review article, or just the introduction to a research paper, you will include citations to the original paper whenever you claim that someone discovered something. It's such a standard practice that the lack of such a reference would be noticed by a referee. — Carl (CBM · talk) 00:05, 23 November 2010 (UTC)

These FA's do not cite the original papers for every discovery and step along the way to mathematical (or physical) enlightenment. However, this guideline requests them to do so: "Where possible, Wikipedia should strive to provide the original reference for any discovery, breakthrough, or novel theoretical development, both for attribution and historical completeness" (my emphasis) Let's look at a section of one of the articles linked much further up: Sylvester's sequence#Applications. For example, "As Galambos & Woeginger (1995) describe, Brown (1979) and Liang (1980) used values derived from Sylvester's sequence", just reads like a literature review. This style of writing, where the author-date bit forms part of the prose rather than a parenthetical citation that could be substituted for a superscript number if a different citation format was used, is typical of literature reviews in academic press. The style exists (as noted above) to give kudos to the authors of the primary research or novel ideas. It is not encyclopaedic. I have no problem with History sections covering notable discoveries and developments (though we should give the full name of such notable researchers if they are indeed notable), and citing the original papers when they do so, in addition to a secondary source that supplies the fact that this is notable (for being first, etc). But the main text of an encyclopaedia should discuss the facts as they are, not tell a story about how we got here and who should be praised along the way.

These FA's do not have an inline citation to five textbooks, one review and one website, as this guideline does (the Annotations section). These FA's do not have a simple basic fact having three separate inline citations, each to a 10-page section in a book, a 6-page section in a book, and to two volumes of a huge and expensive textbook, as this guideline does (the Uncontroversial knowledge section).

What is becoming clear is that this guideline is being used to push a "review of the primary literature" approach to writing WP articles on maths and theoretical sciences. The emphasis is on providing a resource to academics (who are surely well served by professional publications) rather than to the general reader, who I should remind you does not have an honours degree in Mathematics or Computer Science, nor is studying for one.

My biggest problem with this guideline (leaving conflict with policy aside) is both its title and the sentence "This page expands on [ Wikipedia:Citing sources ] and applies it specifically to referencing science and mathematics articles." But "Science" covers a huge range, and I don't think any of this guideline is useful to experimental sciences such as the science behind medicine or biology, say. The purpose of publishing experimental results is to have them discussed, repeated, analysed, rejected, revised, improved upon, etc. For that reason, Wikipedia absolutely must use up-to-date secondary sources and there are grave dangers in using primary experimental research papers as sources for WP (NB: not banned, just limited). Readers will find this guideline and think it is acceptable to write an academic literature review as an article for WP. But reviewing the primary literature is original research. Colin°^Talk 22:08, 23 November 2010 (UTC)

Pillar one reminder

"Wikipedia is an online encyclopedia. It incorporates elements of general and specialized encyclopedias, almanacs, and gazetteers." Geometry guy 23:09, 23 November 2010 (UTC)

On that note I'd like to recall that one of our mottos is "we're collecting the knowledge of the world - including yours" (the exact wording might be off as I'm retranslating that from the German version). That includes specialized knowledge for smaller audiences.--Kmhkmh (talk) 03:11, 24 November 2010 (UTC)

Usual soapbox: WP:Five pillars isn't a policy. It's (just) an essay. Citing 5P as a reason that something must/mustn't be done is like telling the other editor that BRD absolutely requires him to start a discussion every time he reverts something. WhatamIdoing (talk) 21:57, 24 November 2010 (UTC)

Yes, but i think you misunderstand the context. The 5 pillars and the motto are not cited to argue that something must/must not be done, but to motivate why something can be done and more precisely in this case to explain why something has been done already. As I've pointed out further up that we are a collection of special subject encyclopedias as well, is at this point simply a factual observation.--Kmhkmh (talk) 22:10, 24 November 2010 (UTC)

Writing an article on a specialist subject (whether Simpson's episodes, rare fungi, maths concepts, bizarre medical diets, whatever) does not mean writing in a way that only someone with a degree in that subject would either comprehend or appreciate. See policy WP:NOT PAPERS and guideline WP:JARGON. I think this guideline is being used to justify writing postgraduate maths papers rather than encyclopaedia articles. Colin°^Talk 22:25, 24 November 2010 (UTC)

Well yes, as it was repeatedly pointed out to you above there is no disagreement here and people here are aware of those guidelines. Nevertheless there is plenty of specialized material that cannot be described entirely (or at all) in very accessible terms. WP:Jargon btw is mentioning that point more or less literally. And if you read it carefully, then you see that it even assumes that technical articles for "narrow audiences" exist and are unavoidable. The coexistence is no problem, it's just a matter of content organization. There is no problem with providing articles with separate chapters for the more accessible aspects and the more specialized treatments. Note that the fact that we attempt to to provide a greater accessibility for specialized subjects than most normal special subject encyclopedias would do does not not mean we are not a collection of special subject encyclopedias, it just means we are a collection of particularly accessible special subject encyclopedias. Special subject encyclopedias are primarily defined by their scope and content and not their jargon.--Kmhkmh (talk) 23:36, 24 November 2010 (UTC)

The WP:JARGON guideline has all the hallmarks of weasle wording, to be honest. And your "if you read it carefully, then you see that" has all the hallmarks of wikilawyering. The blunt fact is that policy demands we write accessible articles: articles that a non-specialist can read. See my comment below. Colin°^Talk 08:22, 25 November 2010 (UTC)

Nope that's blunt misinterpretation by you. The only thing that policy demands is that we strive for accessibility as far as possible, not more not less.--Kmhkmh (talk) 11:43, 25 November 2010 (UTC)

Whoa, hang on a minute: "WP:Five pillars isn't a policy. It's (just) an essay." and you all agree with that??? That is fabulously wrong on a historical and practical level. The Five Pillars are (and always have been) the fundamental principles of the encyclopedia from which policies and guidelines flow through consensus interpretation of these principles. WP:5 is uber-policy, not an essay! That is something which every editor needs to get straight, whether they wish to wikilawyer or not. Geometry guy 01:26, 25 November 2010 (UTC)

Intuitively and personally I'd see the 5 pillars as a guiding principle or "core policy", but I never payed much attention to technical differences between policies, essays and whatnot and the wikilawyering involved there. I still hope that most stuff still follows from common sense anyway and the 5 pillars are simply common sense from my perspective. Also going simply going by the page design, they are indeed represented as "überpolicy" and there is no indication that they are to be understood as an essay.--Kmhkmh (talk) 01:52, 25 November 2010 (UTC)

Further, before commenting on what a specialist encyclopedia e.g., of mathematics, should look like, it might be a good idea to read one. The Springer Encyclopedia of Mathematics is freely available online. Our articles are considerably more accessible, because we aim to reach as wide an audience as possible. However, that does not mean every article is written at the same level. Geometry guy 02:14, 25 November 2010 (UTC)

It is also interesting to note that we feature specialized math subjects that are not even included in Springer's encyclopedia.--Kmhkmh (talk) 02:58, 25 November 2010 (UTC)

Um, since 5P was written in 2005 (you know, four years after the project started?) I don't think that it's reasonable to describe it as having "always" been anything. For the historical perspective, you might like to take a look at the essay's own talk page archives, paying particular attention to the comments by its original authors that clearly state that 5P is neither a policy nor was ever intended to be one. WhatamIdoing (talk) 04:09, 25 November 2010 (UTC)

"Since 2005" is "always" as far as I am concerned: the project had less than half a million articles then and had barely started. There was a need to articulate the principles of the project and that's what WP:5 does. We post them in our welcome messages, link them first in our "policies and guidelines" templates and refer to them on our policy pages. The original intentions of the authors are no matter: the pillars have outgrown their history and become guiding principles for everything we do, and to refer to them as "just an essay" is ridiculous.

See also "solid hardcore policy" Geometry guy 10:11, 25 November 2010 (UTC)

Discussion at Wikipedia talk:5P already addresses the above. While 5P is an essay, it simply and concisely summarizes policies detailled elsewhere. It introduces nothing that isn't already in a policy elsewhere. If there was something in 5P that significantly disagreed with a real policy, it would be promptly updated to resolve the difference. So let's not get hung up on the policy/essay distinction, it's ephemeral. LeadSongDog come howl! 04:25, 25 November 2010 (UTC)

And regardless of the state of Pillar One, saying that WP has "specialist" content is not the same as saying WP has contents only a "specialist" could read. WP:NOT PAPERS is policy and says "A Wikipedia article should not be presented on the assumption that the reader is well versed in the topic's field." There's no need for wikilawyering. It is straightforward that we cannot assume the reader has a maths degree, or is studying for one or whatever. Colin°^Talk 08:22, 25 November 2010 (UTC)

Exactly, no need for wikilawyering. Maybe you would like to explain how subjects like Pseudorandom generator theorem, Hard-core predicate, Universal hashing, or even Pairing (articles chosen in 10 seconds) could be made understandable to the layman without at least some undergrad maths level and experience in applied mathematics/cryptographic theory. As said above, the general approach is to give an introduction that is accessible as widely as possible to the layman (i.e., as good as possible), but then not to dumb down some subject. Thank you for your understanding. Nageh (talk) 08:34, 25 November 2010 (UTC)

Btw, if you think those articles were not already presented in simple ways I suggest you to look into actual papers on these subjects to not come up with statements implying that people are writing papers here. Nageh (talk) 08:37, 25 November 2010 (UTC)

Maybe you are right that maths papers are even more inpenetrable. The quote on my user page says "It is our job to interest [our readers] in everything. It requires the highest degree of skill and ingenuity." That was made long ago when newspapers had different priorities. I don't think making techical subjects accessible is easy, and ultimately I believe some subjects cannot be covered by WP or cannot be covered in sufficient depth on WP to deserve their own article. But that's my opinion. I think WP should have a discussion on whether it should contain articles where you need a degree in the subject to get beyond the lead paragraph. And that should be on a policy page, not a guideline on sourcing. Currently, I think WP:NOT is fairly clear on the matter, and that causes problems for a lot of our expert editors who want to write expert-level articles. Colin°^Talk 08:57, 25 November 2010 (UTC)

Trying to make the reader interested in a subject is a noble goal, and I second it. However, excluding subjects that you think require too much expertise is clearly not the original goal of an all-encompassing encyclopedia for everyone. I do not understand your resentment towards more special subjects. In fact it is exactly the coverage from trivial topics to specialist topics that make Wikipedia so rich and useful to everyone. And it was ultimately the reason why I have started contributing. I hope these arguments can make you somewhat reconsider your current exclusionist attitude. Nageh (talk) 09:26, 25 November 2010 (UTC)

I agree: "Imagine a world in which every single person on the planet is given free access to the sum of all human knowledge. That's what we're doing." Geometry guy 10:11, 25 November 2010 (UTC)

Well said. Also I'd like to point out that this a voluntary project and we need to work with the authors available to us and we cannot expect them all to possess the "highest degree of skill and ingenuity". If an article is not that accessible or contains too much jargon, then this is usually not an acceptable reason to delete it or not to allow it in the first place, it's merely a reason to improve it (over time). Note that WP a process, that works incrementally for the most part.--Kmhkmh (talk) 11:33, 25 November 2010 (UTC)

That an advertising slogan, not policy. Otherwise, we wouldn't have WP:NOT which excludes most of human knowledge. Please tell me you don't literally believe that line. There's actually only a tiny bit of human knowledge that is encyclopeadic. Anyway, what brought us here was examples of articles that read like literature reviews, which this guideline encouraged. And the desire that our science articles be useful as a bibliography of the primary literature, which is not our purpose. Colin°^Talk 13:02, 25 November 2010 (UTC)

So you are saying that the slogan is a lie, must of human knowledge should be excluded, and you decide what is encyclopedic. I see. Nageh (talk) 13:47, 25 November 2010 (UTC)

Don't be silly. I didn't say "should be excluded". Is said it is excluded. Think about it. Yes, the slogan is "a lie" if you want to put it that way. Very little of "human knowledge" is suitable for an encyclopedia and virtually none of it will ever be written here, even if we have 10 million articles. I think the length of WP:NOT makes that quite clear. Colin°^Talk 15:18, 25 November 2010 (UTC)

no the slogan is not a lie but an strictly speaking an unattainable goal, that we nevertheless to to get close to in a process of approximation. It's life striving for perfection, you'll get get there, nevertheless you spend your whole lifetime to get closer. I believe the slogan illustrates one of the core motivations of the community (or large parts of it) rather nicely and its a foundation for WP's success and attraction.--Kmhkmh (talk) 15:37, 25 November 2010 (UTC)

Have you actually read WP:NOT? It isn't a goal and we aren't ever going to get "close to" it. We aren't a dictionary so we only have articles on nouns and proper nouns. We don't contain original thought that's in your head and mine. We'll never be the first to publish anything, and let's face it most knowledge has never been published. We don't promote or advertise. We don't index the web or provide a directory for anything. We don't instruct people how to do anything. We don't teach people anything. We don't publish novel research. We're not a database of stuff. We don't report the news. We hardly contain any copyright text, pictures or sound. We don't describe non-notable things or non-notable people, which is as close to a definition of everybody as you're going to get. I'm glad the slogan motivates people, but it just a slogan. But hey, this is getting off topic. Colin°^Talk 19:25, 25 November 2010 (UTC)

Hmmm, first "just an essay", now "just a slogan". It has certainly become a slogan, but that doesn't make it a lie. Rather, it expresses an aspiration. Notice that it says "the sum of all human knowledge" not "the sum of all human 'information". There is a difference: knowledge is not indiscriminate information. However, a single word like "knowledge" cannot on its own convey the meaning: you have to understand the aspiration - what do we aspire to make freely available to everyone on the planet? Pillar One provides a pretty strong steer here: encyclopedic knowledge, both specialist and generalist. WP:NOT also clarifies what we mean by knowledge by listing things we don't include. It is a pretty comprehensive list, and it needs to be, because we don't have a WP:IS which lists what we do include. Nothing on WP:NOT comes close to excluding advanced scientific topics or difficult mathematics.

And as WP:PAPER implies, we don't need to restrict the scope of the encyclopedia, because disk space is cheap, so articles aimed at an advanced audience are not produced at the expense of articles aimed at another. I recommend the essay WP:MANYTHINGS for more information on the multi-layered encyclopedia we have built and continue to build. Geometry guy 20:37, 25 November 2010 (UTC)

I'm quite aware of what knowledge is and my point stands. A Wikipedia where experts just write for other experts isn't the encyclopaedia I'm building. I prefer the aspiration in the quote I gave earlier, "It is our job to interest [our readers] in everything.". Call that the sum of human knowledge if you want, but if we can't interest our readers in it then we have failed. Colin°^Talk 22:49, 25 November 2010 (UTC)

Why is it so black and white for you? Writing for a general readership and writing for experts are not mutually exclusive, any more than citation for verifiability and citation for pointers to further reading and citation for proper historical perspective are mutually exclusive. —David Eppstein (talk) 22:54, 25 November 2010 (UTC)

There are topics that we really can't say very much about for a general readership, though, and IMO they still belong here. Take an article I've been working on, scale (descriptive set theory). The first sentence of the article pretty much exhausts what I can say about the topic to someone without a fairly extensive math background.

Nevertheless it is very far from a paper. It treats a concept that comes up in a lot of contexts within descriptive set theory, and that has been used by many authors. The article is a work in progress but is intended to make clear the variety of applications.

Anyone who can think of a way to get more of the gist across to a general readership (without saying anything false, and without getting chatty and vague) is certainly invited to help. --Trovatore (talk) 23:15, 25 November 2010 (UTC)

(edit conflict)There is some group in America, calling themselves Amish, they also think that any technique or scientific advancement that a layman cannot learn and comprehend without expert knowledge has no right to be... something like that. Modern science is an incremental progress, and the times where a single person could attain the universal knowledge of the world are long gone. This means that understanding more advanced topics requires studying the basics first, and whether you are interested in doing so is your decision, but do not make generalizations. Most more advanced maths and science articles on Wikipedia are at undergrad level anyway, and I would hardly call a student an expert. Anyway, you seem to have made up sturdily your mind, so there is nothing we can do to convince you anyway. Nageh (talk) 23:35, 25 November 2010 (UTC)

Wikipedia articles on teenage pop singers don't interest me in the least. Nor do articles on professional wrestling (and they can be pretty incomprehensible). I find some articles on cell biology and medicine hard to understand and I would encourage efforts to make such articles more widely accessible. Would I constrain the encyclopedia only to interest me and write in a way adapted to my understanding? No, don't be daft. Geometry guy 23:06, 25 November 2010 (UTC)

Out of curiosity, how many of the editors here claim to fully comprehend just the lead of Aldol reaction? I don't think math is the only subject that requires mastering the prerequisites. Tijfo098 (talk) 07:42, 26 November 2010 (UTC)

Remaining issues

We've strayed from the point somewhat. The maths experts are writing a different encyclopaedia from the rest of us, and have given up making articles accessible. I can choose to ignore that but when they start imposing their style of writing articles on "science", which the title and the scope sentence includes, then I'm less happy. Really the grand "scientific citation guidelines" is too general and this guideline should be renamed.

The second problem is that this guideline was and to a lesser extent still is badly written and covers stuff that aren't relevant to a sourcing guideline. You guys had ideas about what you were using citations for but the text used to demand we "should strive to provide the original reference for any discovery, breakthrough, or novel theoretical development". That is neither practical or desirable for most of science, and appears to contradict policy at WP:PSTS on sourcing (this is supposed to be a sourcing guideline). The text used to demand "citing the original papers, even if they are not used as sources", which is pretty strange thing for a source citation guideline to advise. The text used to repeat the nonsense about summary style being exempt from WP:V. The text used to have a section on annotations that had a ridiculous and non-specific multi-source citation. These things have been fixed.

The text still has its own "Articles without in-line references" despite the fact that WP has an essay, a guideline and a policy covering this issue quite well thank you. The text spends too much time on eponyms which are nothing to do with source citations. The final section is related to WP:NOR rather than citing sources, so should perhaps go somewhere else. Colin°^Talk 23:28, 25 November 2010 (UTC)

I don't agree with you on restrictions on content, but I kind of do agree about the "original papers" thing. If they're not used as sources, they maybe belong in "Further reading", but not in "References". When you put a general reference that is not used to source a specific claim, that's not always wrong, but it's most likely a general reference — that is, a standard text giving background on the subject, not a research paper. --Trovatore (talk) 23:31, 25 November 2010 (UTC)

"The maths experts are writing a different encyclopaedia from the rest of us, and have given up making articles accessible." Nonsense. Mathematics editors are striving daily to make Wikipedia's articles on difficult topics among the most accessible references on the web (subject to WP:NOTTEXTBOOK). Do you think they are stupid? Please withdraw this opinionated attack! Thanks, Geometry guy 00:33, 26 November 2010 (UTC)

I don't think the complaint is nonsense. In fact, from my own experience, the maths folks are the ones most likely to say that "incomprehensible" articles are just fine like they are, or to respond to complaints about articles being very technical or poorly explained by saying that if you don't understand it, then that's just too bad for you, because any PhD candidate specializing in this particular area of math will certainly grasp it easily.

By contrast, if you complain about the same level of incomprehensibility in any of the applied sciences, you often get at least a minor effort to solve the problem.

I'm not saying that's not the view that all maths editors hold, or even most of them, but I do think it's a somewhat higher proportion than for other parts of the encyclopedia. WhatamIdoing (talk) 02:07, 26 November 2010 (UTC)

Snake lemma as example

If you can manage to make the snake lemma of homology theory (or even Lie group) comprehensible to someone with a reading knowledge of English and no training beyond high-school algebra, fine. If not, consider that the reason that mathematicians say such things is that they have experienced both incomprehensibility and the slow dawn of comprehension. Septentrionalis PMAnderson 06:04, 26 November 2010 (UTC)

I doubt that I could do it for the whole article, but I am convinced that Snake lemma's two-sentence lead could be usefully expanded in ways that would tell readers what it is (more or less), who cares about it, and why anyone cares. Or, for that matter, so that it doesn't link to the same article twice in the first sentence.

But my point isn't about the state of any given article; it's about the typical response, which damages the reputation of maths editors. You said, basically, "can't be done", rather than "I'm sorry that it didn't make sense to you" or "What, specifically, did you find confusing?" or "Let me see if there's at least something I could improve". Your response apparently didn't even include "Maybe I should go read the article to check for vandalism/bad editing/new mistakes"—because I know that your personal editing standards don't support this "department of redundancy department" writing style that the first sentence, and I'm convinced that if you'd actually read the first sentence, you would have noticed it and corrected it. WhatamIdoing (talk) 02:32, 27 November 2010 (UTC)

Agree, even the essay Wikipedia:Make technical articles understandable#Articles that are unavoidably technical requests the lead begin with something the general reader can understand that "attempts at least to put the topic in some broader context" and to "give the lay reader some idea of the place the subject holds in mathematics, what (if anything) it is good for, and what needs to be learned first in order to understand the article". This is also a requirement of WP:LEAD.

I did a search for "snake lemma" and discovered one can buy t-shirts and mugs on the topic, and (as the article notes) it was mentioned in two films! What a shame the article is completely unsourced and (despite the recommendations of this guideline) doesn't tell us who gave it its name. Thank goodness that when the article strays into the advanced and highly specialist reptilian branch of biology that it provides a wikilink. I don't know about you, but my eyes start to glaze over whenever concepts from the real world creep into an article :-). Colin°^Talk 10:19, 27 November 2010 (UTC)

The article is not completely unsourced, but just badly referenced. It contains 2 books, in which you presumably can look up the snake lemma and (hopefully) most of the article's content. It is case of "bad" referencing however, because to the very least the page numbers for those books should have been provided and possibly some inline citations might have been more appropriate.--Kmhkmh (talk) 15:23, 27 November 2010 (UTC)

The fact that the section is titled "Literature" rather than "References" and that they are entire books rather than a page set or chapter means they don't count as references AFAIC. It would be great if someone could fix this. Colin°^Talk 07:53, 28 November 2010 (UTC)

Well that is rather technical distinction regarding section names. A source is a source no matter whether the name of its section is literature,references,notes or whatever. Not giving a page number is indeed bad, but in that context this is most likely not a problem, since well known theorems can be looked up in the table of content or index of those books. So nobody is required to traverse the whole book to verify the information. As I said one needs to distinguish between unsourced and bad referencing techniques here. The article is definitely at least partially sourced, but the reference style/technique needs to be improved.--Kmhkmh (talk) 10:03, 28 November 2010 (UTC)

The discussion is a getting a bit offtrack in particular since we are somewhat arguing nobody really claimed in the first place.

• There is an agreement that all (math) articles should strive for accessibility for the general public. (And various guidelines/policies essays say so)
• There is an agreement that advanced math topics often cannot be described in such a way, that most of the article's content stays accessible to the general public. (And various guidelines/policies essays say so)
• There is (at least almost) an agreement that references to original publication may be used in a complementary fashion, but thre use is not required.
• There was also an agreement that some formulations in this guideline were questionable when the whole ordeal began. (Indeed many of them them have been fixed by now)

There is some disagreement regarding the nature of WP (general encyclopedia only versus combination of general and special subject encyclopedias), but strictly speaking needs to be discussed elsewhere and cannot be resolved here.--Kmhkmh (talk) 15:16, 27 November 2010 (UTC)

A quick google books search shows that Snake lemma is included in a huge number of books. Almost any graduate text on algebra seems to include it. It's also interesting to note that WP:MTAA has been downgraded from a guideline to an essay earlier this year; see discussion, particularly the characterization "strident and uncompromising". Tijfo098 (talk) 03:27, 28 November 2010 (UTC)

Then it should be easy for someone to fix the sourcing issues. Note that Wikipedia:LEAD#Provide_an_accessible_overview and Wikipedia:Manual_of_Style#Technical_language are guidelines and request things that the Snake lemma article doesn't attempt. Note that Tony was referring to this version of the guideline and was part of a drive to slim down the MOS. Colin°^Talk 07:53, 28 November 2010 (UTC)

You are welcome to improve the article. Tijfo098 (talk) 11:17, 28 November 2010 (UTC)

For comparison, the lead of Aldol reaction doesn't explain what the carbonyl "jargon" means except by a wiki link. Do you expect snake lemma to explain every notion that it uses, e.g. abelian category? What's the point of having a wiki then? Mind you, the definition of carbonyl is much shorter than that of abelian category. Tijfo098 (talk) 11:28, 28 November 2010 (UTC)

The lead for sertraline for instance doesn't explain what major depression means (or any of the other disorders that it's prescribed for), nor what tricyclic antidepressants or selective serotonin reuptake inhibitors or what serotonin is or what dopaminergic means (the latter seems to have a pretty ambiguous definition, by the way). Granted some, but not all of these notions have self-explanatory names compared to those in chemistry or math, but only because they relate to everyday experiences (panic, anxiety, stress, etc.) I'd be curious to know how many readers without a priori knowledge of cognitive behavioral therapy can correctly infer what it means from its name. Also, I'm willing to wager that more than half the infobox for sertraline is completely meaningless to the average reader. Tijfo098 (talk) 11:53, 28 November 2010 (UTC)

Even articles like B-major assume that the reader knows what a major scale is and what a key signature is. Someone who is entirely unfamiliar with music theory will not be able to understand much if anything there. I completely agree that the idea that math articles are uniquely incomprehensible is a myth. — Carl (CBM · talk) 12:43, 28 November 2010 (UTC)

Actually, some of this confusion may be caused by the polysemantic word "jargon", So I've started a discussion about that section at WT:MOS. Tijfo098 (talk) 00:41, 29 November 2010 (UTC)

Other issues

This section isn't covering any "remaining" issues; it's just a reiteration of the same issues from the other sections above. Tijfo098 (talk) 07:16, 26 November 2010 (UTC)

Yes, but it's true that the unfortunate digression over whether advanced topics are suitable for Wikipedia at all (my view, lest anyone doubt it, is of course they are) has obscured one or two fairly good points put forward by Colin. The idea that we should strain to use original sources strikes me as pretty silly. Gödel's work, for example, is (almost) universally adjudged brilliant; his seminal paper, On Formally Undecidable Propositions..., is on the other hand very far from a good source for his theorems, which have since been much polished. In cases like that one, we'll mention the paper anyway, just for its historical value, but for less revolutionary work I hardly see the need. --Trovatore (talk) 07:35, 26 November 2010 (UTC)

I agree that a wording that says original works should be cited even when not used as sources does not make sense. I do think that the guidelines should state that citing original works is desirable for scientific breakthroughs and novel theoretic advancements, both for the significant value of such papers, historical accuracy, and attribution. This may be done as easily as in "In the seminal paper by XYZ,^[ref] ...". Primary sources in such cases are meant as complements to secondary sources and not as replacements. Exceptions may be made for (1) very recent papers, which are peer-reviewed but for which secondary literature does not exist yet, (2) highly specialized results, as long as they are peer-reviewed and cited by other scientific papers, and (3) standards documents that are issued by officially recognized standards organizations such as the ISO, IEEE, or IETF. Please, think about why I excluded these three categories before crying out again. Nageh (talk) 09:41, 26 November 2010 (UTC)

I've explained my take on the issue of what should be cited and for what purpose in #Why it is important to cite original sources above. Tijfo098 (talk) 21:41, 26 November 2010 (UTC)

The special role of mathematics/formal sciences

I want to address this non-stopping complaint by some people that maths people are ignorant and do not want to make their articles comprehensible. This is not the case. The problem lies somewhere else. In contrast to all other sciences, mathematics and similar formal sciences (computer theory (in Knuth's sense), logic, statistics) do not rely on observations in real life, i.e., they are no empirical studies. This makes is somewhat difficult to explain advanced topics to laymen because for empiric studies you can always say "this follows because that can be observed", and everyone will understand, even if not comprehending the deeper connections. In mathematics, you would say "this follows because of that theorem", but while an observation is easily accepted as the cause, a theorem is usually not understandable at all to a layman. In empirical studies you may ask "why can you observe that?", and you may follow a wiki link to another topic explaining the phenomena. Doing so is limited in mathematics because you cannot insert (meaningful) wiki links in mathematical expressions, and while you may add an explanatory remark saying which rule(s) you have applied, as long as the rule is no specific (named) theorem the endless chain of rules that had to be noted to go down to the axiomatic system makes this effort impossible. From this follows that understanding mathematical expressions requires acquiring a certain understanding of the mathematical language. This is in principle no different than learning other (natural) languages where you need to start with the spelling, the grammar, and simple vocabulary to comprehend more advanced texts. The difference though is that while spelling and vocabulary (the mathematical notations) and grammar (the axiomatic system) are quite small in the mathematical language, the increasingly complex phrases (theorems) that you can build based on these rules is sheer overwhelming, and the best you can do to understand more advanced texts is to learn and understand a lot of these phrases (as well as techniques to construct new phrases).

Another aspect makes explaining formal sciences often difficult to the layman. Formal sciences are often inherently theoretic, with no direct (but certainly many indirect) applications in real-life. This makes writing a layman's introduction difficult, because the immediate use is only for other mathematical theorems.

I am sorry if I have not been able to express myself more clearly on this subject. You may want to inquire your nearest philosophy professor, s/he will certainly be more capable.

In conclusion, I think that the least of the maths folks are as ignorant as some may think, as it is ultimately the desire of anyone (any teacher) explaining a topic to make it as easily accessible as possible. If you think we are not doing our best you are welcome to make constructive suggestions for improvement.

Thank you for your reading, Nageh (talk) 09:11, 26 November 2010 (UTC)

Further reading: the source for Computer science#Paradigms, [2] Tijfo098 (talk) 21:34, 26 November 2010 (UTC)

Apology

Thank-you for your explanation, Nageh. I don't think you or anyone else here is ignorant, indeed erudite is the word that comes to mind. I've been on Wikipedia for five years, and (apart from a period when I was active at Featured Lists) have spent what little free time I have for WP trying to ensure our medical articles are accurate and accessible. Our medical experts (of which I am not one) are, I believe, particularly willing to target their articles at the "general reader" because they feel little need to write for their peers. What doctor would rely on Wikipedia for medical advice, or medical student for information for her exam, or researcher for a review of the literature? They have their own professional publications and online resources. Another reason many medical experts write on Wikipedia is to ensure the Web has an accurate resource on the subject, when there is so much crap elsewhere. Mathematics is not unique in having topics where the barrier to entry is high: biochemistry, molecular biology and genetics can be impenetrable in areas, and medicine has plenty jargon to contend with.

Like many passionate debates, this one has got polarised. This has the unfortunate effect of making one write in a more "black and white" (as someone put it above) way. It also has led me to make assumptions based on what I've read here rather than perhaps what is true if I were to read more articles and contributions. At the centre has been this imperfect guideline. I'm grateful to those editors who have overcome any (natural and understandable) defensiveness and have agreed the wording could be improved. Misunderstanding is the cause of many an argument: there's what the text said; there's what I interpreted it as saying; and there's what you thought it said or assumed it said.

I have held that Wikipedia is written for the "general reader" for so many years that it is difficult for me to accept differently. I've read the WP:MANYTHINGS essay that Geometry guy linked, and it may surprise you that I find myself agreeing with much of it. The lack of editorial control and resource limitations (WP:NOTPAPER) isn't necessarily a good thing, however, when one wants to write an article that won't bore the reader rigid. I was taught in writing that the first step is to determine who one's audience is. Another vital skill is knowing what not to say just as much as what to say. Perhaps it is unfortunate that Wikipedia is the only successful wiki knowledge-base, and that professionals and academics don't have their own space (Medical alternatives like Ganfyd and Medpedia haven't had our success, I don't know if the maths folk have tried anything). Or perhaps that's a good thing as it encourages experts here.

I have respect for those who wish to share their knowledge on Wikipedia and am sorry if I've made any of you feel bad. I haven't "made up sturdily [my] mind" and this discussion has certainly given me something to think about. I hope we can all be friends on Wikipedia. Sorry this response is a bit long. I give you one of my favourite sketches: Armstrong and Miller Physics Special You've probably seen it before and I've posted it before elsewhere. Rather appropriate, I think. Still makes me laugh. Cheers, Colin°^Talk 22:36, 26 November 2010 (UTC)

No hard feelings. Even a good discussion can get lively and emotional at times. Thanks for the response, and the link, it was amusing! :) Nageh (talk) 18:24, 27 November 2010 (UTC)

Well spoken. I would just respond to one point. I don't understand what benefit you think there is in moving inherently difficult material to another resource besides Wikipedia. Where do you draw the line, and why?

The links between articles can easily take you bit by bit across very different "intended audiences". Number has a link to the rather more difficult, but still basic, notion of a real number. Somewhere in that article, it is no doubt mentioned that the reals have a larger cardinality than that of the naturals. From there you may want to read about the continuum hypothesis. Recent arguments addressing CH's truth or falsity could lead you to Ω-logic, one exposition of which makes reference to the Stone–Čech compactification.

The last article in the list, at least, is in the category of things about which very little can be explained to a reader without a very substantial background. But why should it be cut off from the others, as an external rather than internal link? It's going to have a lot of links back to less forbidding topics. Or maybe you would cut Ω-logic off, even though we can at least tell the general reader that it's a strong way of making inferences that can't be deduced in ordinary logic?

I just don't see the advantage, unless it's to keep people from hitting articles that they don't understand. But why is that a problem? None of us understands everything. "Write for your audience" is fine, but "audience" should be understood as per-article, not per-encyclopedia. --Trovatore (talk) 22:01, 27 November 2010 (UTC)

I'm still thinking about the "per-encyclopedia" aspect and so would rather not respond on that. As for "per-article" audience, that's a traditional publishing/writing concept but WP seems to have adopted a different approach. Where an article is irredeemably complex, it should have at least some of the lead accessible (see comments above re Snake lemma), and where a section of an article is similarly too complex, it should also have some kind of lead sentence/paragraph that at least lets the general reader take something away. Colin°^Talk 23:05, 27 November 2010 (UTC)

There is no disagreement about better leads. Nobody argued against providing a better more accessible lead.

The (potential) disagreement arises if you argue that such topics should not be covered WP at all or that they need to be deleted if the lead is not optimal with regards to accessibility.

Also, there is some sense of priorities needed here. The first priority is to have a correct and sourced article on a particular subject at all. Then as a second priority, we can focus on providing optimal accessibility. Also in light of "per article" argument above, you could state a rule of thumb, that the more specialized a subject is, the less important the accessibility for the general audience becomes. That does not mean that we don't want accessible leads for those articles, it just means it is not our top priority, i.e. we have usually more important things to worry about. We need to keep in mind that WP is collaborative effort with a big diversity in skill sets and interests among contributors and that WP and its article generation is a process. One author may write an a good and sourced article for a specialized audience and another author might add a more accessible leads and some more accessible aspects later on (or vice versa). We need to work with diversity of our contributors and not against it.--Kmhkmh (talk) 11:46, 28 November 2010 (UTC)

Actually, the idea that the lead could be made more accessible is being argued against, in the Snake lemma section above. And I'm being given a WP:SOFIXIT response, which is rather ironic given that the issue is that I haven't a clue about the topic. Plus some silly arguments that wafting a maths textbook at the article makes it sourced. We do have a naming scheme for article sections, and nowhere is "Literature" recognised as the source section. Try suggesting a general textbook is a "source" for a single point of knowledge at GA/FA and see how much that attitude would get appreciated. I'm seriously considering unwatching now because the defensiveness, unwillingness to engage and desire to argue rather than to admit something isn't as it should be, isn't making for a useful conversation, and certainly isn't helping Wikipedia. Colin°^Talk 16:46, 28 November 2010 (UTC)

No offense, but if you need a section header to be able to identify a source and incapable or unwilling to understand how a textbook can be used as a source, then indeed it might be best if you refrain from commenting here.--Kmhkmh (talk) 17:01, 28 November 2010 (UTC)

You can't just make a rude comment and stick "no offense" in front of it. Guidelines are supposed to be about, you know, best practice. There are very important reasons why we put specific citations in specific sections: because it is the promise from the editor to the reader that says "I read this and it verifies the text I wrote". Citations elsewhere or entire-book citations are just for "further reading" and don't have same contract. But this is all too draining. Rather than constructively discussing how an article or guideline might be improved, I get insulted for being blind and thick. </unwatching> Colin°^Talk 17:14, 28 November 2010 (UTC)

No offense is manner of speech. And yes if you constantly ignore the arguments made with somewhat rude innuendo yourself ("defensiveness", unwillingness") then at some point you will get a rude reply. Guideline and policies are not only about best practices but also about minimal standards and required practices. There is nothing wrong with having standard sections, but there is something with wrong expecting every author that writes an article to be aware of all them and making a big fuzz over a "false" section title. And let me repeat again what I said further up "completely unsourced" is not the same "inappropriately or insufficiently referenced". What matters ultimately that the given sources allow readers and other editors to verify the information in the article and those 2 books under the former "literature" section did that at least partially. Also in the discussion further you were again framing the issue in false manner. Neither me or anybody suggested the referencing of the snake lemma was already appropriate, in fact I said explicitly further up that it wasn't. What I did object to was of your claim that the article was completely unsourced, since it wasn't. Whether an article needs inline citations or not depends on the particular article (its length, how many sources it uses, whether are quotes or disputed content) and the author has to make judgment call on that. The snake lemma is probably a borderline case, meaning for (science) articles of that size only depending on 1-3 short sources covering the content it may be sufficient to simply list those sources at the end.--Kmhkmh (talk) 18:39, 28 November 2010 (UTC)

Let's not be bureaucratic here. We do have naming conventions, but even they acknowledge that there is no univocal agreement on section titles for sources. Regarding the accessibility of article lead sections, it is a good concept and I think most (all) of us agree that it should be encouraged. Trying to make a point on immature articles is pointless, though.

What concerns deleting (seemingly) inaccessible articles, I think that discussion is somewhat over for the moment. Nageh (talk) 17:30, 28 November 2010 (UTC)

exacrly--Kmhkmh (talk) 18:39, 28 November 2010 (UTC)

Heads-up

There's a new essay and proposed guideline on the contents of "Further reading" sections at WP:Further reading. It is still only a couple of days old, so constructive comments are more useful than !votes at this point. I'm posting this notification because it's related to the extensive discussion here of when to use inline citations, the advice given here on further reading annotations, and the discussion when the original/primary sources are worth mentioning that way. Tijfo098 (talk) 04:42, 30 November 2010 (UTC)

Very helpful and reassuring and inaccessible

I am new to Wikepedia and I had a coding problem that has been resolved and a stylistic problem that has not. My experience may suggest an opportunity for additional cross referencing that might save future users from despondency or departure.

The cross referencing problem

I posted a request for help on my User talk:Michael P. Barnett page to find out how to make multiple references to a single citation. I received quickly a very clear reply routing me to Wikipedia:Referencing for beginners. This resolved my coding problem, but not the styling difficulty. I am not certain how I got there, but I think I decided that if there is Referencing for beginners site there might be a Referencing site. There is, and that is where the round of troubles that I have just resolved began. My transfer to the Reference site was redirected to Citation. Scanning this, by eye, led me to Scientific citations. I looked at the main article briefly, then at the Discussion, and commented (agreeing) with the two comments that had been posted. Then I looked at the article itself, and typed a detailed critique. Essentially, it seems to belong to a body of information (?) that is disjoint from the professional world in which I have worked for over 60 years. During the two weeks in which I have edited Wikepedia articles, the only comments I have heard from high school and college students and their parents, and librarians, are that Wikepedia is very unreliable and must NOT be cited in any papers as a source of information. I was not encouraged by Scientific citations.

I found the site where I am typing in a web search! I have tried several link jumps from Citation and they lead to Scientific citations. Neither of the categories Bibliography or Reference lead here.

Some questions that can be answered very briefly

• Might it be advantageous to put some more links in Citations and do something about the mismatch between the present Scientific citations site and the professional world that I think most working scientists live in?

• Might it be advantageous to give a link from the site where I am typing to sites that are optimal for users outside the fields you specify?

• Is the specification of these fields unduly restrictive? Doesn't this site cover astronomy (and space science in general), geology (and earth science in general) and a lot more of biology than you mention (anatomy, physiology, for starters).

• Might it be helpful to include, early on this site, some comment on how Wikepedia style differs (if it does) from the style of the thousands of journals that now accept (request) LaTeX files coded using the \cite command.

The stylistic question

I have worked with base aligned reference numbers for so long I have difficulty restructuring my sentences to superscript style, without the result looking bizarre. For example, how do I rewrite: "The principles that underlie the methodology are explained, e.g., in [1, 2], the theoretical basis in [2, 3, 4], and actual examples of its applications comprise or are included in [1,4,5]". This may just be a flaw in my writing style that I have to amend to work with Wikepedia, but I find myself writing sentences like this very often. Thanks Michael P. Barnett (talk) 02:47, 6 December 2010 (UTC)

Try, "The principles that underlie the methodology are explained, e.g., by Smith^[1], Jones^[2], the theoretical basis in Jones^[2], Johnson^[3], and Black^[4], and actual examples of its applications comprise or are included in Smith^[1], Black^[4], and White^[5]." Ozob (talk) 11:53, 6 December 2010 (UTC)

Really grateful. I was still locked into the mind set of publication costs (dating back to 1951). Your message set me off wondering how much was being saved. The paraphrase increases the character count from 190 to 251. Each of the 8 references in the paraphrase contains a single name, with average length 5. For the corresponding paraphrase with an average of 2 authors joined by " and " per paper (just a guess) and average surname length 6 (based on 50 commonest English surnames), the increase would be 136, i.e. about 70% in storage needs, and somewhat less on the screen and in a printout. So I will adjust to the times. Michael P. Barnett (talk) 15:46, 6 December 2010 (UTC)

P.S. Would it be potentially helpful to future users to include, under heading "Writing for superscript citations": Authors who are accustomed to the use of baseline aligned citation numbers may need to alter their writing style slightly to use superscript citation numbers. For example "The principles that underlie the methodology are explained, e.g., in [1, 2], the theoretical basis in [2, 3, 4], and actual examples of its applications comprise or are included in [1,4,5]." can be paraphrased to "The principles that underlie the methodology are explained, e.g., by Smith^[1], Jones^[2], the theoretical basis in Jones^[2], Johnson^[3], and Black^[4], and actual examples of its applications comprise or are included in Smith^[1], Black^[4], and White^[5].". If it is worth putting this in, should it go here or in Wikipedia:Referencing for beginners or both? Michael P. Barnett (talk) 16:46, 6 December 2010 (UTC)

My feeling is that it would go better in Referencing for beginners.

Also, here's another idea: Harvard references. As in The principles that underlie the methodology are explained, e.g., by (Smith 2003), (Jones 2005)" and so on. There are templates for this such as {{harv}}. Whether or not Harvard references look funny depends on what you're used to; in my own field of mathematics they're rare, but in the social sciences I hear they're everywhere. Ozob (talk) 01:53, 7 December 2010 (UTC)

Humanities, too: It's the official style required by most university English departments, for example. The information is at WP:PAREN. I think it is particularly valuable for non-technical articles that I expect to get a lot of attention from new users, since the referencing system works just like it did when typewriters were the hot new technology. WhatamIdoing (talk) 18:40, 31 December 2010 (UTC)

Math references (if done in the usual LaTeX/BibTeX way) are somewhere between parenthetical references and footnotes. I never write references with footnotes (some journals disallow footnotes even for parenthetical comments). But the references in math journals usually look like "[2]" or "[SMH2009a]" instead of "(Smith, Martin, and Han 2009)" so they aren't quite "Harvard style" either. I find that I am more comfortable with parenthetical references on Wikipedia because they seem closer to the math ones (especially the [SMH2009] kind) and because they are easier to read without looking at the bottom of the article. — Carl (CBM · talk) 18:47, 31 December 2010 (UTC)

Related discussion

Editors here may be interested in the discussion at Wikipedia talk:Citing sources/example style#why_not_standardize_on_one_format.3F, an effort to impose a one-size-fits-all citation style on every article. My overall impression is that the goals are to ban general references, to require the use of <ref> tags (banning WP:PAREN and all other forms of WP:Inline citations), and to require the use of citation templates. WhatamIdoing (talk) 04:20, 11 January 2011 (UTC)

Page numbers and abbreviation

Since Wikipedia is not paper, we rightly advise editors that journal titles need not be abbreviated, even though that is common practice in printed journals. I would like to suggest that we extend the same advice to page numbers. Many articles use the unpleasant PubMed approach of abbreviating page numbers when they are over 100. Not only is this unnecessary, it is also potentially confusing: "110–99" (meaning "110–199") could be interpreted as a backwards range of 11 pages, rather than a forward range of 89. That confusion is not very likely, I admit, but since there is no reason at all for using such abbreviations, could we not add a note to that effect to this guideline? --Stemonitis (talk) 08:54, 13 February 2011 (UTC)

Given the wide range of articles that might follow this guide, I think that advice like that is too localized. I also use the "full" form (100-123) but if some other field usually uses the short form I don't think we should try to force them to change. — Carl (CBM · talk) 13:16, 14 February 2011 (UTC)

This guideline (and it is only a guideline) applies specifically to scientific citations, as does my suggestion. What other fields were you thinking of? Since Wikipedia is not paper, we don't have to follow minutely the established traditions in publishing; indeed, we don't do that, because abbreviating journal titles is absolutely standard. Can anyone supply another reason for abbreviating page numbers? --Stemonitis (talk) 13:33, 14 February 2011 (UTC)

We don't have to follow tradition to save paper, but that doesn't mean we have to always break with it. This guideline has never focused much on the "fine" details of citations. It does suggest giving the full journal title, and including info like the ISBN or DOI, but it doesn't go into fine detail on how to actually format the reference.

I would be worried about starting to focus on the "fine" presentational details like how to format page ranges, whether to say "pp." or not, etc. I think those can be left to the article-by-article decision process. (I also don't think many people do use the 100-11 format, in any case.) — Carl (CBM · talk) 13:39, 14 February 2011 (UTC)

"provide ... an original example"

I think this is potentially controversial. It probably needs more clarification as to how original the example may be. I propose that it's okay to provide an original example when it can be easily verified by the reader given the information in the article, and assuming fluency at level of the topic discussed. If I came up with Stone duality myself (and did not publish it elsewhere first) and gave that as example at equivalence of categories, or something derived from that (like Stone's representation theorem for Boolean algebras) as example at isomorphism, doing so would be pushing the boundaries of WP:OR way too much. Tijfo098 (talk) 22:28, 18 March 2011 (UTC)

The question of "how much is too much" is probably unsolvable in general; we deal with in pretty well on individual articles, though. WP:TECHNICAL also mentions the point that examples are useful for conveying technical information. I removed the word "original" from this guideline, since it can only cause headaches. But I think that trying to be too precise is trying to churn water into butter. — Carl (CBM · talk) 23:27, 18 March 2011 (UTC)

Seems fine to me. Tijfo098 (talk) 01:40, 19 March 2011 (UTC)

ref to Manifold Destiny?

Right after that sentence there is a footnote saying "See Manifold Destiny for a possible counterexample." I miss the subtlety, and I suspect I'm not alone in this... Tijfo098 (talk) 03:50, 19 March 2011 (UTC)

"Attribution" section title is a poor choice

This section is isn't a discussion attribution in the sense of WP:NPOV, but rather when to cite wp:primary sources for something having implicit attribution, like a named theorem, which (evidently) is already attributed. Tijfo098 (talk) 22:41, 18 March 2011 (UTC)

Isn't "attribution" the normal word for attributing results to their discoverer? — Carl (CBM · talk) 23:23, 18 March 2011 (UTC)

Surely it is, but the thrust of that section is not ensure attribution, which is already done by the secondary sources (us naming theorems would probably be WP:OR), but rather to provide complementary citations to primary sources (something that irked fokes in natural sciences a great deal in the sections above because they didn't understand the epistemic exception of formal sciences in that respect). Or at at least that's how I read it. Perhaps there are two issues? The section starts with "Some statements require attribution." The first example is "Supernova 1987A was discovered by Ian Shelton and Oscar Duhalde at..." That statement is not attributed in the sense of WP:ATTRIBUTEPOV, but rather written in "Wikipedia's voice" as a matter of incontrovertible fact. Attribution in that example would be "Textbook author X reports that Supernova 1987A was discovered by Ian Shelton ...", which would be silly. Tijfo098 (talk) 01:07, 19 March 2011 (UTC)

I suggest renaming to "Citing landmark primary sources" or something like that, which is the thrust of the section, and dropping the confusing 1st sentence. Tijfo098 (talk) 01:14, 19 March 2011 (UTC)

I'm saying that in the non-Wikipedia world, "attribution of results" means mentioning who proved a theorem first, who did an experiment first, etc. It is not the same as "providing a citation", it's the act of ascribing something to its creator, a kind of name-check. This is a normal English meaning of "attribution". So you're right this is not related to WP:NPOV, it's about giving credit where it's due. — Carl (CBM · talk) 01:26, 19 March 2011 (UTC)

Sure, that's true, but this is a Wikipedia guideline, so the question here is: what does that have to do with citing the primary sources, which is what the section's text is about? Surely, I can cite a textbook just as well to say X's Theorem is named after X. My point is that the section's title is uninformative as to its contents. If come here pondering whether to cite a primary source or a textbook for a theorem (and WP:PSTS/WP:RS generally steers people towards the latter), I'd have no clue that I should look in the section called attribution for advice on that. That's why I'm saying the section title needs some wiki-keywords like "primary sources" in it. Tijfo098 (talk) 01:40, 19 March 2011 (UTC)

"Attribution" was the title of the failed merge proposal for WP:V and NOR. A person speaking plain English, rather than wikijargon, would read that and guess that you were talking about WP:INTEXT attribution. WhatamIdoing (talk) 02:54, 19 March 2011 (UTC)

My point is the text in that section is not about in-text attribution. Just look at the last example there... (In-reference attribution?) That's why I say the section title is confusing. Tijfo098 (talk) 03:26, 19 March 2011 (UTC)

I edited the first paragraph some; does that make it more clear? — Carl (CBM · talk) 11:58, 19 March 2011 (UTC)

Yes it's better now. Thanks, Tijfo098 (talk) 14:49, 20 March 2011 (UTC)

Last example from that section even more troublesome

If I read it correctly, it's an experimental result from a primary source. What if another source comes up with a different number? In any case it's confusing at best for anyone not knowing the topic, so it probably doesn't belong here. Tijfo098 (talk) 02:02, 19 March 2011 (UTC)

I don't like this example for a different reason: there is no in-text attribution of the authors, so it's not an example of attribution. I'd say we can remove it. — Carl (CBM · talk) 11:58, 19 March 2011 (UTC)

I don't see it as a different reason. I guess I didn't express myself clearly. What I was considering was: what if some other authors find a different number/bound? Then you need some sort of in-text attribution to distinguish the two. Sometimes it's done in a less transparent way, without naming the sources. ("A meta-analysis in 200x found this, but a later meta-analysis in 200y found the opposite.") This happens a fair bit in natural science articles, particularly in GWAS studies these days. Tijfo098 (talk) 14:24, 20 March 2011 (UTC)

Full names, or initials?

In regard of author and editor names, should we lean towards full names (as in "Smith, John"), or last name and initials ("Smith, J.")? - J. Johnson (JJ) (talk) 18:40, 27 August 2011 (UTC)

There's a third alternative to these two options: citing the names as they appear in the original publication, regardless of whether that leads to spelling the same person's name two different ways in different citations within the same article. Anyway, I don't have a strong opinion about this, but I think there are significant differences in conventions between different areas of science, so if the guideline says anything about it it should be that all three of these conventions are ok, that we should choose a consistent convention within each article, and that we shouldn't gratuitously change the convention that has already been chosen for an article. —David Eppstein (talk) 19:35, 27 August 2011 (UTC)

There is much to be said for conforming to the form of the initial publication. But then we likely have inconsistency within an article.

The question arises because I am developing a canonical format for the principle IPCC publications. And within various articles there is no consistency to be consistent with (such conventions not deliberately chosen, but developing ad hoc), so I feel free to proceed in any direction. But which way? - J. Johnson (JJ) (talk) 19:00, 28 August 2011 (UTC)

See WP:Centralized_discussion/Citation_discussion#Full_name, or initials? for additional comments.

- J. Johnson (JJ) (talk) 22:33, 30 August 2011 (UTC)

In the context of this page, it's worth remembering that the vast majority of medicine-related journal articles pull their citation data from PubMed, which is initials-only. WhatamIdoing (talk) 00:21, 31 August 2011 (UTC)

Query raised at Original research noticeboard

I have raised a query at:

Wikipedia:No original research/Noticeboard#Can new formula be put in to a scientific article?

about an issue at Planck's law and the interpretation of Wikipedia:Scientific citation guidelines#Examples, derivations and restatements Dmcq (talk) 12:44, 2 November 2011 (UTC)

Routine calculations: median?

I've just deleted the following phrase, from the recently added list of examples of routine calcualations: summarizing data sets (e.g., by providing the range, mean, or median rather than the full list of raw data). There has been some controversy over this issue at Usage share of operating systems. Certainly finding the median of a list of numbers is a simple calculation, but the interpretation of the result is unclear. When summarising a data set via a median (or mean or other statistic), it's possible to create a misleading impression; the assertion that a single number is a valid summary of a data set is dangerously close to original research in my opinion. I think that a clear consensus needs to emerge before adding any guidelines to this page. Jowa fan (talk) 00:47, 18 November 2011 (UTC)

Like the range and the mean, the median should not be used for any sort of interpretation, but merely presented as a simple and basically indisputable fact. You don't say "the median height of US Presidents is ___, which means that..."; you simply say that the median height is ____, and leave any interpretations up to the reader.

The goal here is to provide a description rather than raw data, not to advance an argument. WhatamIdoing (talk) 02:42, 18 November 2011 (UTC)

I agree that this makes sense in some contexts. But it's clear from the extensive discussion at Talk:Usage_share_of_operating_systems#Summary_of_issues_with_median_row_of_the_usage_share_table that there isn't yet a consensus on this issue. Jowa fan (talk) 05:12, 18 November 2011 (UTC)

Actually, what I hear in that discussion is that one user was unhappy about it in a single article and that Dmcq is opposed to every single application of CALC that I've ever seen him comment on. The principal opponent, by the way, voluntarily proposed retaining the median at the Mediation Cabal case yesterday, which amounts to an admission that finding the middle number in a list of numbers is not actually a policy violation. WhatamIdoing (talk) 05:46, 18 November 2011 (UTC)

Harv templates nominated for deletion

There's been a proposal to delete the harvard citation style templates. I know the detailed citation style is not really the subject of this guideline, but editors here may still find this of interest. Please see Wikipedia:Templates for discussion/Log/2012 January 24#Template:Harvard citation and comment there. —David Eppstein (talk) 00:22, 25 January 2012 (UTC)

Proposed new cleanup tags

I've encountered a fair bit of confusion over how the policies on primary sources apply to scientific citations, since the situation is different than it is in non-scientific fields. Specifically, the scientific guidelines support the use of both secondary sources such as review articles as well as primary research articles. I would like to propose the following two cleanup tags as an alternative to {{Primary sources}} for scientific articles:

Please let me know what you think. Antony–22 (^talk⁄_contribs) 02:13, 22 February 2012 (UTC)

The first one seems useful and I would use that - a lack of review articles is very common. The second template seems less useful to me, because I rarely request that editors try to find more primary sources. Blue Rasberry (talk) 13:54, 1 April 2013 (UTC)

The existence of review articles is very field-dependent. In experimental fields, where there are many papers with original data, review papers are more common; in theoretical fields that are not data-based, review articles are much less common. So there should not be a general expectation that review articles exist for each article topic here. — Carl (CBM · talk) 13:58, 1 April 2013 (UTC)

Citation unease

I am having arguments with science article editors about the usefulness of citations. Some seem to think that citations are not needed for assertions. On Talk:Big Bang I am having a cn tag reverted because an assertion "does not need a citation". A similar argument is going on in Talk:Uncertainty principle because "it is egregiously unreasonable to expect a truckload of one's preferred quantum mechanics texts tacked on for "verification"?!" I fear that there are a great many of these editors, which accounts for the lack of citations in science articles. Myrvin (talk) 15:20, 24 January 2015 (UTC)

I concur with these concerns. Le Prof (User:Leprof_7272). 71.201.62.200 (talk) 21:19, 6 June 2015 (UTC)

Weak citations

I am coming across science articles where an editor uses a whole textbook as a citation. Perhaps this guideline should stress that, for books, the page number(s) in which the support resides needs to be stated. An example of this is in Uncertainty principle, where I am having difficulty arguing that more citations are needed. An editor has now replaced my tags with a citation for a whole textbook. His/her argument is that the article is not a tutorial. Myrvin (talk) 13:30, 23 January 2015 (UTC)

"That person is what I like to call 'wrong'." (Source:Some tv commercial I once saw.) You're looking for the {{pn}} tag. LeadSongDog come howl! 17:31, 23 January 2015 (UTC)

Found it thanks. Myrvin (talk) 17:37, 23 January 2015 (UTC)

It is very occasionally acceptable to use a whole textbook without page numbers as a citation: when the article is so short that it hasn't been divided into subtopics or footnoted and the whole textbook is itself about the same subject as the article. But I think as soon as you reach the point where you need to be providing separate inline citations for distinct claims within the article, you have also reached the point where you need to provide page numbers in the inline citations. (It's still acceptable for those inline citations to point to an entry in the references section that doesn't include the page numbers, for instance if you are going to have multiple footnotes with different page numbers all pointing to the same book, of course). I agree with LeadSongDog that {{pn}} is the right way to tag these bad citations. —David Eppstein (talk) 19:11, 23 January 2015 (UTC)

And in a case where you're only using one source, you can always use {{rp}} as a way to reference individual pages without creating separate reference entries. Guettarda (talk) 19:29, 23 January 2015 (UTC)

I prefer to put the page numbers into the actual footnotes but {{rp}} is also an acceptable citation style. —David Eppstein (talk) 19:38, 23 January 2015 (UTC)

If it's all the same, yes. I was thinking about the case where you write an entire article mostly from one textbook, and you reference multiple groups of pages in a single book. In a case like that, it's a useful alternative. Guettarda (talk) 22:12, 23 January 2015 (UTC)

Oh please don't use {{rp}} -- it's just too bizzare, even ugly, and tends to confuse the readers. It's a weak sop for those who can't figure out how to do short cites. ~ J. Johnson (JJ) (talk) 22:25, 23 January 2015 (UTC)

Disagree based on WP policy, sorry. the "rp" manner of adding pages and avoiding redundant citations is an allowed style of WP, is used in science articles (where the short citation style is less the norm), and editors should continue with the styles that appear in an article until discussion with interested editors makes clear a decision a uniform, overall change in style should be undertaken. I won't cite WP: this and that, here, but this is what WP policies state. 71.201.62.200 (talk) 22:20, 6 June 2015 (UTC)

Short cites? What should I be doing instead? (I used it because it seemed to be the least confusing, least ugly option I could come up with, actually.) Guettarda (talk) 22:36, 23 January 2015 (UTC)

{{sfnp|Author|Year|pages=xxx}} —David Eppstein (talk) 22:55, 23 January 2015 (UTC)

Oh, those things. That's the problem. If you want to know what ^[1] refers to, you have to go down to the reference section. That disrupts the flow of reading, but a back button will take you back to where you were. But when you hit something as cryptic of "Smith (1992), p.2-17" you need to take a second step to see what Smith (1992) is. And this one you have to do on your own, scrolling down past the other references to see what it is. Then you need to find the place where you were.

To the average reader, the source matters more than the page number. The page number only matters if you want to look things up. Granted, {{rp}} doesn't eliminate that problem, but at least it keeps the source just a click away. (Can you tell my background is not in a field where they use footnotes and ibid very often?) Guettarda (talk) 16:00, 24 January 2015 (UTC)

Thank you everyone, but the question on the table is can we beef up this guideline to say that citations to books should always contain the page numbers on which the support can be found? How about something in Citation format" along the lines of: "For cited books, it is important that the pages, in which the supportive evidence can be found, are included." Myrvin (talk) 16:51, 24 January 2015 (UTC)

Sometimes what is useful is not necessarily a page number, but could be a section number, or even a paragraph number. In general, these are the specification of the specific location within the source where the material is to be found.

I disagree with Guettarda that for "the average reader, the source matters more than the page number." I think it is less a matter of more importance than that for a first-level validation of sources it is often sufficient to know just the source (typically author[s] and year). If the reader is familiar with the field, or has already looked up the source, the additional bibliographic detail is not necessary. (If not, then the additional details of title, publisher, etc., are quite likely not sufficient.) Having an additional detail (e.g., page number) is no hinderance, but not having it greatly hinders actual verification. And as our requirement here is, per WP:V, "looking things up", not the reader experience, I am strongly inclined to require specification. ~ J. Johnson (JJ) (talk) 23:52, 3 February 2015 (UTC)

The style of citation should simply follow established practice in the article, per wp:CITESTYLE. Provision of the information necessary to wp:V is the sine qua non of Wikipedia. Whether page numbers are needed or not is something of a judgement call based on size and structure of the source: is it easy to find the relevant bit within it? LeadSongDog come howl! 02:56, 4 February 2015 (UTC)

WP:V says: "Cite the source clearly and precisely (specifying page, section, or such divisions as may be appropriate)." Perhaps we could have something like this in the guideline. The danger is that there are some editors who think that page numbers are not appropriate - or even a citation at all. 20:10, 4 February 2015 (UTC)Myrvin (talk)

I can only disagree with LSD, and concur with Myrvin as he states the WP:V policy. Moreover, the ambiguity is less than indicated; e.g., the instruction for use of states a page range of "over two pages" (see [3]). Page or other narrowing numbers (section, etc.), always, see more below. Le Prof (User:Leprof_7272). 71.201.62.200 (talk) 22:20, 6 June 2015 (UTC)

That is indeed a problem: some editors are antipathetic to page (etc.) specification. In some cases I suspect they actually want to discourage verification, because they know they're on thin ice. More often I think it's more a matter of not wanting to take the trouble. So while I agree with LSD that page (etc.) numbers are not always necessary, I think we do need to encourage them. ~ J. Johnson (JJ) (talk) 22:22, 4 February 2015 (UTC)

I disagree entirely regarding the lack of an absolute need, early in an article's history, to establish that page numbers are always included when referencing texts—because of the poor pattern it sets that later entering editors (esp. inexperienced ones) will follow, and because the secondary point of sourcing is to provide places for those being educated to turn for further material, and page number omission hinders this secondary aim.

In the former case, the trajectory is clear: an article that begins without good sourcing always is more likely to develop and continue to have poor sourcing (akin to the earliest pattern). In the latter case, it is fallacious to assume that a reader or student can find their way to the actual source content within a book (hence, the WP policies stating that even long articles should have a narrowing of page ranges); to presume the least of readers can do what the original writer/editor was lax to do—this is to overestimate the ability of readers, and to release editors to editorial sloppiness.

For an example of the mess that develops if the wrong patterns develop early, see the Human sexuality article, and the deep, pervasive issues with whole book citations there. See also my comments below on the existing policy text here, in two specific cases, offered as a Prof, until recently at a major research university, working at the interface between the physical and biological sciences. Le Prof (User:Leprof_7272) 71.201.62.200 (talk) 21:53, 6 June 2015 (UTC)

Citation policy clarifications

The reasoning in sections of this article/article is often circular, and the examples poorly persuasive; they appear to be clearly written by persons with a very great distance from the needs and expectations of a typical WP user.

The three physics statements here, Wikipedia:Scientific_citation_guidelines#Uncontroversial_knowledge, should never appear without a citation: they are not common knowledge in any sense (stated as a physical sciences prof), and they further demand, for sake of readers, at least one source each that would allow a non-expert reader to followup the all-but-gibberish, jargonistic content to a source they could actually read and understand. The existence of such "policies" here bespeaks WP's early origins in rapid content generation from communities with histories of online content creation, where the level of "communication to oneself" was staggering.

But WP is not a maths or physics forum: it is a general, authoritative encyclopedia. If a statement is not understood to the high school and early collegiate reading population, prima facie, it is over-technical, and without sources appearing to allow education to continue, over-technical content appearing in Wikipedia is generally useless. It only continues the time-honoured process of academics talking to themselves (before classrooms, in their writings, etc.).

The very same applies to the example of the Aldol reaction—a reaction I have performed and taught, and article I admire and whose principle academic author I have met, but an article that is not as precisely sourced as it needs to be to be generally useful to younger students. (You will see, if you look to discussions of this very good article, that issue has been taken with its level being too sophisticated; this concern is exacerbated when large tracts of text lack reference to more remedial explanatory sources.)

Finally, the reference to the low basis theorem article, here, Wikipedia:Scientific_citation_guidelines#Articles_without_in-line_references, as an example of proper sourcing is all the more egregious. One should never allow stubs to be accepted as articles without proper inline citations. Let us put on our thinking caps. With 1-2 sentences, and all sources in bullets, yes, the correspondence between content and source is clear. But what transpires when immediately following edits are done, and further content is added? Two things, clearly:

• First, the correspondence between the original material and its sources is diluted, and eventually lost.
• Second, the arriving editors are more likely to follow the existing pattern, and add further material without inline citation, simply adding their source to the bulleted list, and while diluting the initial material-source correspondence, completely obscuring it for the newly added material.

Moreover, half the citations in the earlier and current low basis theorem article are books without page numbers, and all are either books lacking page numbers, or long articles of 30-40 pages or more. No, a thousand times no. This is not a good example of how we should direct editors to create new content for science articles.

Strikethrough added based on the focused clarification offered below, by Mr Eppstein. The bulk of the argument clearly remains unchanged, see below. Le Prof 71.201.62.200 (talk) 23:43, 6 June 2015 (UTC)

On both of these points, these policies are achieving what they inadvertently support: sloppy sourcing practices on new and developing articles. These practices give tremendous support to views that WP articles are simply not useful as materials through which students learn about science and its communication (except, as here, as sources of poor examples in pedagogy). Le Prof (User:Leprof_7272). 71.201.62.200 (talk) 21:19, 6 June 2015 (UTC)

It was false that the book references in low basis theorem omitted the page numbers: they were present in the parenthetical referencing in the text of the article. However, I have also added them to the references section, so maybe next time they will not be as easy to miss. —David Eppstein (talk) 22:37, 6 June 2015 (UTC)

First, this—may I suggest smug?—clarification of page numbering relates only to an improvement in the page numbering in recent versions of the article, but the clarification is irrelevant to the actual version cited by the existing policy (where the page numbers were and remain absent as of this time/date-stamp, see [4]). Second, the changes you made address a superficial aspect of the point being made—your change from an uncommon style (year-colon-page number) to a common style (page numbering appearing following "p." in the citation) is indeed helpful—but ignores the general point, that the article originally, and for some time lacked this, supporting the need for a general demand that page numbers always appear from the onset.

Third and finally, the cited article remains a very poor example of how to completely and adequately cite maths or any other science articles. Insofar as we are now able to trace the material in the text to the precise points in the citations the editor intended, it simply does not help a general reader very much. For a mathematician speaking to other mathematicians, the job is done; for one seeking to make the appearing material less gibberish, to any but those already deeply understanding the subject—that aim has not yet begun. Even the limited depth of the clarifying prose provided to other mathematicians, in Jockusch, Jr and Soare's original paper:

"Using the methods of recursive function theory we derive several results about the degrees of solvability of members of certain… classes of functions (i.e. degrees of branches of certain recursive trees). As a special case we obtain information on the degrees of consistent extensions of axiomatizable theories, in particular effectively inseparable theories such as Peano arithmetic, P."

which is ripe with terms that might be wikilinked (or otherwise defined by citation), even this degree of reader-directed clarification is omitted from this article. Hence, more is expected of our readers than of the readers of these expert's original work. My criticism of this as representative of good sourcing for a general article intended for non-specialist readers of an online encyclopedia remains in force. Le Prof (User:Leprof_7272). 71.201.62.200 (talk) 23:38, 6 June 2015 (UTC)

Derivations versus proofs, and the definition of "original research"

In mathematics, proofs and the techniques used in proofs are often as useful, if not more useful than the results themselves. Under item 4, "Examples, derivations and restatements" it says that a including a different derivation is perfectly acceptable, and in fact "encouraged", but this leaves open the possibility that the only available reference for a proof using the technique used in a Wikipedia article is the article itself.

For example, in Atom (measure theory) there appears a theorem on non-atomic measures. The proof that appears on that page uses Zorn's lemma. The proof by Wacław Sierpiński appearing in the cited article does not. The closest I could find, using the Wikipedia page as a starting point, is the hint in exercise 2:14.8 (a) of Real Analysis that says to use "some form of Zorn’s lemma." If I am writing a mathematical paper, exactly what am I supposed to cite as the source of the proof given on the Wikipedia page, the hint in the exercise in the source document, the Wikipedia page itself, or "obviousness", or "folklore"? In what way is the proof given in the Wikipedia article not original research?

I am placing my question here and not on the talk page of that article because I believe that the question is addressed to Wikipedia guidelines that apply to more than just that one page. --Penguian (talk) 12:24, 25 June 2015 (UTC)