Wikipedia talk:Scientific citation guidelines/Archive 1

From Wikipedia, the free encyclopedia

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Contents 1 Why must the "references" be so terse? 2 Referencing discoveries and breakthroughs 3 Two more 4 footnotes and more 5 status 2006-10-13 6 Comments 7 Proposal or guideline? 8 When not to use in-line references 9 Further discussion needed 10 Project guidelines? 11 Disputed 12 Response to Sandy's criticisms 13 Minor issues about eponyms 14 DOI numbers? 15 must versus should 16 not just math and physics 17 Echo Sandy 18 Thanks, this is helpful 19 Concerns 20 Question on relationship to GA/FAC 21 annotations creep 22 Consensus among math and physics editors 23 New MathSciNet template 24 Endorsement From The MCB Project 25 Link to journals in references 26 where to link and why 27 articles about journals 28 Good articles - Space weathering 29 Endorsement from Wikipedia:WikiProject Chemistry

Why must the "references" be so terse?

I agree with the reasoning expressed in this page. However, I find the suggestions less helpful than they might be.

The natural sciences have their own citation conventions. They work for the natural sciences; I don't want to criticize them, or the way journals may require their use.

But whatever else it may be, WP isn't a journal in the natural sciences. There's no requirement for a single list of endnotes-cum-references, let alone for the endnotes, references, or combination to be as terse as is entirely suitable within a journal. As the proposal now stands, the only description I can see within the references is "(Review)". I can guess that this means it is a review of research by others rather than a presentation of original research, but I don't even know that -- and even if my guess is correct, I don't know if the research is very recent or if these are reconsiderations of a decade's worth of work.

At what is now the end of a slightly heated exchange here, I suggest a way to meet the requirements for "verifiability" and also to suggest further reading for what is widely (or even universally) accepted within mainstream physics but is sadly little known (and legitimately questioned) elsewhere. I'm candid about my appalling level of understanding of physics (well, I did manage to get my physics A-level at the second attempt) but I think I'm polite and coherent. Autoplagiarism time: If the content of an article can be verified in any of a great number of sound, contemporary physics books:

you do either or conceivably both of two things: (i) have a single discursive note near the top of the article (perhaps at the end of the first paragraph) saying that the ideas summarized in the article are argued for in one, two or three books; (ii) list these two or three books in "References" together with brief descriptions of their relationships to the article.

I elaborate on this using a non-physics analogy. Of course this can and perhaps should be supplemented by specific citations for certain specific points.

Nobody has responded to my suggestion. Maybe it's just too stupid, maybe what I wrote is just too boring.

For an example of an article in which notes are discursive and are separated from references, see Pierre Rossier. Meanwhile, Hiroh Kikai adds descriptive notes to the references.

(Yes, these two have considerable similarities to each other, have much more notes than I imagine are needed for physics articles, and are about matters that are stunningly irrelevant to physics. I'm sure there are much more helpful examples; when I think of them, I'll add them.) -- Hoary 03:54, 2 October 2006 (UTC)

I support (ii) over (i). In addition, cite requests should not be used by people who'd like a second opinion because they distrust wikipedia - they should be used when one has an informed cause to disagree with something that wikipedia says. –M^T 05:07, 2 October 2006 (UTC)

The fact is that articles get delisted as Good Articles or are denied as Featured Articles not only because they are considered not well-referenced, but also specifically because they do not have "sufficient" inline cites, where "sufficient" is not defined. Based on articles like Lancaster County, Pennsylvania it may be feared to get inflated to several per sentence. Apparently you are not a member of the inline-citation squad. Still, it should not be required to cite things that are common knowledge in a field. It should not be necessary, each time something is said of the nature that fish live in the water, that Helium is a gas, or that Carbon has atomic mass 12, to give a reference for that. Part of what makes some reasonable people so upset, next to the intransigent stance of the verification police, is that it is illusory to think that citations create verifiability for people who do not understand the basic concepts of a field. If you do not understand the basics, you are not in a position to judge whether the presentation in an article corresponds to the sources, and also not whether the selection of sources is reasonable. Most people just cannot distinguish pseudoscience from science. The requirement to provide citations that serve no reasonable function does not increase the quality of Wikipedia. It makes articles look unprofessional if not ridiculous, and is further only annoying.  --Lambiam^Talk 05:35, 2 October 2006 (UTC)

Apparently you are not a member of the inline-citation squad. No, I'm not a member of the squad. I've submitted three or four articles for GA, of which a couple have passed. I've passed at least one to which I'd made no significant contribution, I've expressed doubts about two or three others. GA seems a good idea severely damaged by factors independent of inline citations, and perhaps also by a demand for inline citations. I'm not so interested in "Good Article" status; I'm much more interested in the quality (accuracy, informativeness, helpfulness) of articles. ¶ If you do not understand the basics, you are not in a position to judge whether the presentation in an article corresponds to the sources, and also not whether the selection of sources is reasonable. Very true. And I doubt that anyone will be able to grasp the basics even from excellently written encyclopedia articles. People rarely acquire a knowledge of new (to them) and difficult subject areas from encyclopedias or other reference works; rather, they learn from coursebooks or introductory works. And that's one reason why I suggest pointing toward one or two of these books. ¶ Most people just cannot distinguish pseudoscience from science. I think this assertion conflates two things that can and should be distinguished. The smattering of theoretical physics that percolates through the broadsheet press, etc., to the ignorant (I'm one) looks so screwy and counterintuitive that no, we indeed can't distinguish between it and the occasionally pseudointuititive concepts of pseudophysics. However, the moderately well educated and moderately intelligent (I like to think I'm one of these, too) can distinguish between a pseudoscientific and a scientific approach. We know that there's a world of difference between, say, Newtonian physics and "creation science". (Elsewhere, we may know little about efficacious psychotherapy, but we're appalled by the chameleon claims made for quack psychotherapies, etc.) ¶ Yes, it's unreasonable for us physics ignoramuses to demand (easily digestible!) "verification" for everything. But it's not unreasonable if we refuse simply to take the article editors' word for everything. Tell us where we may get a fuller understanding of (and perhaps also a certain degree of verification of) the claims that are made in the article. -- Hoary 06:56, 2 October 2006 (UTC)

But hopefully you agree that not everything has to be cited from sources, such as the factoids that fish are water-dwellers and come in different sizes. The inline citation squad would require our Fish article to go like "A fish is a water-dwelling vertebrate with gills, that remains so throughout its life.^[1]." Do you feel (whether inline or not) that there should be a source referenced in the article where the reader can actually find this information on page so-and-so? Note that even this is not good enough for the verification police. They require explicit pointers to the location within the source, such as page numbers. Further, it is one thing if this is sourced in the main article on fish, but should we have a separate citation for each and every time a statement in some article implies that fish live in water? What about the roughly one thousand times the atomic weight of some element is mentioned? I think we all agree that some things must be cited. Hopefully we also agree that not everything must be cited everywhere, and that, in fact, such a practice would be a disservice to the readers and editors alike. So the question is: can we draw up guidelines that say when to cite and when not to cite -- with a grey area in between that is left to the editor's discretion? For academic subjects such guidelines are best developed in consensus between the editors knowledgeable regarding that subject, and should reflect de facto standards and good practice for the field. The present proposal is an attempt at that. If there is a reasonable guideline, people may actually follow it. If the bar is unreasonably high editors will ignore it, or, if forced to follow it, just drop out.  --Lambiam^Talk 12:17, 2 October 2006 (UTC)

But hopefully you agree that not everything has to be cited from sources, such as the factoids that fish are water-dwellers and come in different sizes. Yes of course I agree. ¶ The inline citation squad would require.... [abridged]. I'm not part of any such squad; I'm not much interested in any such squad. ¶ I think we all agree that some things must be cited. Hopefully we also agree that not everything must be cited everywhere, and that, in fact, such a practice would be a disservice to the readers and editors alike. Yes, fine. ¶ So the question is: can we draw up guidelines that say when to cite and when not to cite -- with a grey area in between that is left to the editor's discretion? For academic subjects such guidelines are best developed in consensus between the editors knowledgeable regarding that subject, and should reflect de facto standards and good practice for the field. The present proposal is an attempt at that. Then I think it's a little misguided. The most useful citation practice within a good general-purpose encyclopedia (whose readers aren't limited to students and scholars on academic subject X) for articles on X is very likely to differ from conventional and convenient practice within X. ¶ If there is a reasonable guideline, people may actually follow it. If the bar is unreasonably high editors will ignore it, or, if forced to follow it, just drop out. Or jump to the gruesomely named but yet intermittently promising Citizendium. ¶ The first couple of footnote-reference hybrids on the project page are interesting: they're to what look as if they might be accessible books. But this forced hybrid of note and reference (which I think is normal within the natural sciences, where it may be very useful) seems to prevent useful elaboration, such as (imagined example) "An accessible account of carbon-carbon bonds in general, suitable for those with a good grasp of introductory chemistry" or (again) "A technical account, for advanced undergraduate students of chemistry and above". -- Hoary 13:13, 2 October 2006 (UTC)

---

I think your comment about annotations is quite reasonable. I had tried to do something similar on the physical cosmology page some time ago. I added a little section about that to the article. Hopefully nobody disagrees too stridently with it. I don't know how much annotation we'll see. This page is meant to present guidelines, not hard and fast rules, and I think the convention within physics is usually to give the references as succinctly as possible: "Mitchell, R. J. et al. 1976 M.N.R.A.S 176, 29." It is rare that you see combined footnotes and a references section in physics: usually either you have the references in the footnotes, or do Harvard referencing with purely textual footnotes. It is unusal to have seperate footnotes and references, although I agree in many cases it would be desirable in Wikipedia. –Joke 16:22, 2 October 2006 (UTC)

Yes, the references in physical cosmology look very helpful indeed. (A minor tip, though: take a quick look at what's now the very first "external link".) The obvious question (from certain factions) is: "All right then: there's a lot of stuff in this article and you're listing a lot of books and links; so just what comes from where?" And I'd agree that this question would be a rather stupid one. Still, there's a nugget of reasonableness in it. One or two prefatory sentences might help. Imagined example: "All or virtually all of the content of this article is explained in greater depth in any of the following works, all of which are rather technical and assume considerable ability in mathematics." ¶ Another point: Though I'm coming more from the humanities and have appeared to criticize the terseness of citations in the natural sciences, temperamentally I'm all for terseness and I find it ridiculous that otherwise sane adults expect each other to have rainforests hacked down in order to provide space for such horrors as "Princeton, N.J.: Princeton University Press" (as if there were other, confusible Princeton UPs in Queenston, N.M., Pusan, S. Korea, etc.). -- Hoary 23:20, 2 October 2006 (UTC)

Referencing discoveries and breakthroughs

Beyond eponymous theories and laws and numerical data, any discovery or breakthrough in general that can be credited to an individual or group should be cited as such. --ScienceApologist 20:43, 4 October 2006 (UTC)

Agreed. (Although in some cases, it might be very hard to find.) –Joke 01:48, 5 October 2006 (UTC)

Hi, this is the irritating non-scientist piping up again. I note that the first in the list of sample eponyms that should be credited is the Michelson Morley experiment. In certain contexts* it may be good to explain in a note (or otherwise) precisely which paper is being referred to; but in general wouldn't it be better to just to provide an internal link to Michelson-Morley experiment?

* Notably of course within the article Michelson-Morley experiment.

Hoary 09:58, 5 October 2006 (UTC)

The draft includes However, articles that link to these may choose not to cite the original papers, depending on the context. In this case, a reader looking for the reference may easily click the article link to find it.. This could be clarified to point out that a wikilink to the article should be included if the reference is not. I think that some editors will want to link to both the wiki article and the original paper.CMummert 12:29, 5 October 2006 (UTC)

Well said. Sorry for my carelessless. -- Hoary 12:38, 5 October 2006 (UTC)

Two more

I think this proposal is a good start towards writing down the practices that well sourced technical articles should follow. Please do announce it to the math project, since it seems suitable to us as well as to the physics project.

Here are two points that are not currently covered:

• Derivations (or trivial restatements of theorems) that are not original research but may not literally appear in any convenient source. Example: Planck's_law_of_black_body_radiation#Derivation or Decider#Relationship_to_partial_Turing_machines.
• Is Harvard referencing acceptable? All the examples use footnotes.

CMummert 23:25, 4 October 2006 (UTC)

Added them both in. I didn't know what in your examples didn't appear in a convenient source, so I used my own example. (Don't tell me that the simple analysis I did on Lambda-CDM model doesn't work because the errors are not independent and non-Gaussian. The errors on h and Ω_bh² are approximately independent, and close enough to Gaussian that I don't worry there will be a New York Times article bringing down Wikipedia because of it.) –Joke 01:48, 5 October 2006 (UTC)

Thanks. Although I don't know a reference (who writes about conventions of science authors?), I agree that examples of basic concepts are not considered original research when included in articles or textbooks, and so should not be considered original research here if consensus is that the results are truly basic consequences of the definitions. If they are self contained then they are trivially verifiable. CMummert 02:23, 5 October 2006 (UTC)

Some people do write about science writing conventions, actually. Here is a relevant quote from Steven G. Krantz, A Primer of Mathematical Writing, AMS 1997, p. 77 (re inline citing vs a section of bibliographic references at the end of a chapter, though not really addressing the question we're discussing here of what level of statement needs a citation):

"These little end-of-chapter essays can be quite informative and, if well written, can give the reader a sense of the historical flow of thought that in-context references do not. I would say that the down side of this end-of-chapter approach is the following. It serves the big shots well. If you are annotating a chapter on singular integrals, then you will certainly not overlook Calderón, Zygmund, Stein, and the other major figures. But you might overlook the smaller contributors. The advantage of the in-text, on-the-fly reference method is that it systematically holds you accountable: you state a theorem, and you are given the reference; you recall an idea, and you give the reference. You are much less likely to give someone short shrift if you adhere to this more pedestrian methodology."

—David Eppstein 20:29, 5 October 2006 (UTC)

footnotes and more

The proposal currently claims footnotes are common in physics and math journals. I don't know about physics journals, but the math journals I use typically have Harvard style references (example: "Smith [2000] was the first to propose this idea") and only have footnotes for long parenthetical comments. I don't know the relative frequency of footnote citation vs. Harvard style citation, but the claim in the proposal seems strong to me. CMummert 12:38, 6 October 2006 (UTC)

Yes, the statement was becoming increasingly tenuous so I removed it. I believe that in Physics all the Phys. Rev. journals as well as the Elsevier journals like Nucl. Phys. and the online journals like JCAP use footnotes, but a number don't, such as the Astrophysical J. and (I believe) one of Nature and Science. So best to remove it altogether. –Joke 14:36, 6 October 2006 (UTC)

The proposal says

Likewise, if an editor requests that a particular statement be sourced, it may be advisable to add an in-line reference. This may be provided on the article's associated talk page, although it is worth adding it to the article if it would prevent confusion.

I think a reference on a talk page should not be called an inline reference. Also, should the spelling inline be used everywhere? in-line is used a few times in the current draft.

My concern is that several people on recent WP:CITE discussions seemed confused between citations in general (which are required by WP:V) and inline citations, which are required by the GA group but not by WP:V. CMummert 12:42, 6 October 2006 (UTC)

Yes. Another confusion people make is the different usage between "citation" and "reference." I wasn't meaning to imply that a talk page comment was an in-line reference, but the way I phrased it certainly seems to imply that. –Joke 14:36, 6 October 2006 (UTC)

The proposal says

In general, Wikipedia should strive to provide the original reference for any discovery, breakthrough, or novel theoretical development, both for attribution and historical completeness:

I agree. I think we should add another sentence for clarification:

When the original reference is not suitable as an introduction to the idea, either because it is outdated or because it contains serious errors, an annotation may be added to the citation noting this fact.

Many mathematics original references will be both outdated and contain errors, and so they are not the place a general reader should look to learn about the idea. CMummert 12:50, 6 October 2006 (UTC)

It's much the same with physics. Done. –Joke 14:36, 6 October 2006 (UTC)

What we also need are inline citations like "[1]" which appears on the same line as the text instead of like a superscript like footnotes. Most physics journals use this citation style. In wiki articles the citation causes an extra gap between subsequent lines, which looks ugly. Count Iblis 14:47, 6 October 2006 (UTC)

It is probably possible to add something to monobook.js (is that what it is called?) that does this for you. I'm not sure. –Joke 14:53, 6 October 2006 (UTC)

Put

sup.reference { vertical-align: baseline; }

in your monobook.css and the inline reference superscripts will no longer be superscripts. It works for me. CMummert 15:04, 6 October 2006 (UTC)

That will screw up exponents in many math articles. —David Eppstein 15:12, 6 October 2006 (UTC)

No, it only affects sups that are of class "reference" which does not include the sups that wikimedia creates for math. You can see the difference if you view the html source code for this: x^{2 [1]}. I see the superscript 2 but the reference is not raised.

Also, changing your own monobook.css has no effect on anyone else; it is a way to change how WP looks to you. So for example I have mine set so articles are 7" wide even if my browser is wider. It's a matter of personal taste. CMummert 15:34, 6 October 2006 (UTC)

Thanks, it works great! I've moved the citations up by 0.1em to allign them a bit better with the text: sup.reference { vertical-align: baseline; position: relative; bottom: 0.1em;} Count Iblis 15:56, 6 October 2006 (UTC)

You're right; sorry for the confusion. Now if only CSS could make it as easy to switch between footnotes and Harvard style... —David Eppstein 16:29, 6 October 2006 (UTC)

It won't be long before people start running BibTeX on Wikipedia. I've often thought about doing it myself. –Joke 17:28, 6 October 2006 (UTC)

BibTeX is not very different from Wikipedia:Citation templates. One drawback of the <ref> ... </ref> version of inline citation, btw, is that it makes it difficult to read and edit the source if templates are used (the reference grows to too many source characters and it interrupts the flow), while untemplatized references are often inconsistently formatted. That's why I tend to use Harvard style when I'm editing an article that doesn't already use the footnote style. —David Eppstein 17:42, 6 October 2006 (UTC)

Yes, except that BibTeX is better than citation templates because it lets you keep a database of references apart from the text of your article (and it doesn't screw up the formatting, unlike the citation templates). Plus, unlike the ad-hoc templates used here, it is completely standardized. I often contemplate writing a wikipedia.bst style for BibTeX to dump references to a file in either citation template format or pure Wikipedia markup. That way I could copy and paste my BibTeX references directly into Wikipedia. Perhaps someone has done it already? –Joke 19:22, 6 October 2006 (UTC)

status 2006-10-13

What is the status of the proposal right now? I noticed it has been archived from the physics project talk page. Assuming the text here is stable, maybe now is a good time to put reminders on the appropriate project talk pages, and see if there is consensus to adopt this as a project policy in physics and/or math (and/or other science projects). CMummert 22:29, 13 October 2006 (UTC)

I think so. Maybe as a direct link from WP:CITE. --ScienceApologist 15:49, 14 October 2006 (UTC)

I thought a link from the per-project guidelines would be good for the time being. The only reason to change WP:CITE is for GA review, but I don't think the GA reviewers will care about a per-project guideline. CMummert 16:38, 14 October 2006 (UTC)

I agree. I don't have time to do it this weekend, but I'll do it Monday if nobody gets to it first. –Joke 21:24, 14 October 2006 (UTC)

For the longer term the more prudent thing to do is to follow process and aim for a status as "sub" guideline having community support, even if only targeting science articles. But I guess it wouldn't hurt to "beta test" it first amongst ourselves.  --Lambiam^Talk 23:53, 14 October 2006 (UTC)

Well, this would be a subguideline in the sense that all the wikipedia policies still apply, and probably take precedence. The point of it is to give more specific guidance than the general policies give. I spent a while discussing WP:CITE on its talk page, and found no willingness there to include any specific citation guidelines on that page. But I think there is a long-term need to have some standards about what should be cited and what does not need to be cited. If there is evidence that several scientific projects agree with a specific set of citation guidelines, that would give us a stronger argument when arguing for a more rational policy for wikipedia overall. CMummert 01:14, 15 October 2006 (UTC)

I bet if we included a link from WP:CITE that nearly no one would object. --ScienceApologist 01:22, 15 October 2006 (UTC)

Comments

You asked for comments on Wikipedia talk:WikiProject Mathematics, and they seemed best here. For the name Wikipedia:Scientific citation guidelines seems the best rather than making it "owned" by a particular project. The writing is still rather defensive, if we are thinking of a policy or guideline we should avoid too much hedging, which will encourage criticism. It might be good to expand on the standard practice of citations in scientific literature, some examples of standard citation practice in papers might help get other up to speed. Derivative is a good mathematical example, where almost the entire article is very standard stuff. Indeed some of the statements there can be verified by the reader from first principles. It might be possibly be worth mentioning history sections, these tend to benefit from inline cites as this sort of material is less well know in the literature. --Salix alba (talk) 23:45, 16 October 2006 (UTC)

Another guy from maths. I agree about the defensive part (assuming that Salix means what I think) and I edited the lead section to address this. Feel free to edt / revert / shred to pieces / send to BJAODN. It would be good to have a quote from a style guide that one doesn't need to provide a citation for standard stuff. On the name, I remember that there is an editor arguing on the talk page of WP:CITE (or thereabouts) that there are far more inline cites in medicine. Obviously, we shouldn't go for Wikipedia:Scientific citation guidelines unless we have consensus in all the sciences; perhaps Wikipedia:Citation guidelines for the physical sciences is an easier goal. On the aldol reaction example: Why are refs 1-3 and 4-6 in two seperate groups? The references should be perfect, so include the title and full journal names. What do you think about mentioning explicitly in the footnote that these are standard references for verifying the article (this is basically the point that Hoary makes above)? Finally, on a strategic level, we should keep an eye on the proposed Wikipedia:Attribution which may lead to some shifts in Wikipedia policy. -- Jitse Niesen (talk) 14:37, 21 October 2006 (UTC)

I generally support it, but I find the statement "articles that link to [eponymous articles] may choose not to cite the original papers, depending on the context" too vague. I would prefer if such cases were handled just like links from a summmary to a sub-article. This would reduce the "dense referencing" and facilitate maintenance, since the sub-article is the best place to discuss and maintain attribution. — Sebastian (talk) 05:35, 19 October 2006 (UTC) (Copied from Wikipedia_talk:WikiProject_Mathematics by Joke 16:47, 23 October 2006 (UTC))

Hi, I'm sorry I didn't see these comments until now. As far as it being rather defensive, yes, I agree that that is true. It was written in this way partly to emphasize that it is, in fact, a guideline and that it is meant to interpret and apply WP:V, WP:CITE and WP:NOR rather than supercede them, and also to avoid including any explicit statements like "don't provide a citation for this..." which would almost certainly cause someone to object. Moreover, I quite agree that the examples could certainly be improved. Most of those I included are from articles that I have worked on, or that I am at least aware of, which makes them rather lopsided. I will try to fix up their formatting and journal names, etc, today.

As for the point Jitse made about annotations (and that Hoary made above), I thought this was now covered in the section "Annotations"?

As for the case of eponymous articles, I should probably solve the problem with an explicit example. If, in an article about the aether, you said "...the Michelson-Morley experiment showed that the speed of light is the same in all directions..." then it shouldn't be necessary to provide a citation. However, a case where it would be good practice to provide a citation is when something is not likely to be discussed in the linked article, like "The first use of platinum mirrors in interferometry was the famous Michelson-Morley experiment.^{[citation needed]}" (n.b. this statement is not, to the best of my knowledge, true or even sensible.) I will try to write something to reflect this.

Finally, Jitse expressed some worry that the very vocal propoents of inline citations wouldn't see that these guidelines are reasonable. I agree that there is always going to be someone who isn't pleased, but I think I can infer from the response I've seen from outside the WikiProjects that most people will accept that they are a sensible compromise. –Joke 16:47, 23 October 2006 (UTC)

Proposal or guideline?

After reviewing the edit history, which reveals only a handful of editors reviewing this proposal, I switched it from guideline to proposal. The title "scientific" is too broad, since it appears so far to encompass only a few members of the math and physics projects. Sandy (Talk) 14:22, 26 October 2006 (UTC)

The title is not too broad. It has been submitted on the main pages of both WikiProjects, has the consensus of those projects, and I have stated that I would now like to ask the members of other related projects if they endorse the guidelines. I should note that many proposals have quite broad names, see Category:Wikipedia proposals. I reverted your change, but I accept that something other than the standard guideline check mark is more appropriate in the banner, so I used the proposal question mark. –Joke 14:42, 26 October 2006 (UTC)

Sandy has alerted me to this discussion. I'm inclined to agree that it should be treated as a proposal at this stage. Tinkering with WP's referencing requirements should be done only with the utmost caution, since the authority of the project is at issue. Tony 14:49, 26 October 2006 (UTC)

Hi. I see Sandy suggests that these are "weaker" citation requirements or that this proposal somehow changes the basic primacy of WP:CITE or even WP:V. I think if you read over the guidelines, you'll find that they are intended as a set of clarifications about how the verifiability policy and cite guidelines ought to be implemented rather than a weakening of the requirements. The main point, I think, is to clarify the difference between referencing an article about a very advanced topic like cosmic inflation – which I have recently done quite a lot of work on, and I doubt (aside from a few citations I haven't dug up yet) anyone would say is poorly referenced – and an article like derivative (mathematics), which is about quite widely known material in thousands of textbooks, and how everything in between can be handled. –Joke 14:58, 26 October 2006 (UTC)

Yep, I think the standard is to mark it as a proposal first until consensus has been sought and it deems it a guideline. --plange 15:13, 26 October 2006 (UTC)

(edit conflict) The guidelines clearly represent a consensus in the maths and physics projects. It was proposed some time ago and nobody spoke out against it, as far as I know. The text makes it clear that it only applies to physics and maths at the moment. I don't know how the edit history tells how many have reviewed the article. Of course, I'm very interested in opinions on the contents of the guidelines / proposed guidelines. -- Jitse Niesen (talk) 15:24, 26 October 2006 (UTC)

If the "proposal" represents math and physics, but has not yet included consultation with broader scientific projects, it is labeled too broadly, and should not be considered a "scientific" "guideline". If it applies, so far, to math and physics, that is a relatively small part of all "scientific" articles. Sandy (Talk) 15:41, 26 October 2006 (UTC)

So your problem is the title, you want it moved to, say, Wikipedia:Citation guidelines for physics and mathematics, and move it back if it is accepted for scientific articles? I guess you have a point. I'm not sure it's worth the effort, but I won't oppose it. -- Jitse Niesen (talk) 16:22, 26 October 2006 (UTC)

Either something like that, or just label it a proposal in the interim (that would be easier than a move, no?). I have no specific preference: just that both "science" and "guideline" are currently too broad. Sandy (Talk) 16:25, 26 October 2006 (UTC)

When not to use in-line references

On this section, it might be beneficial to state that these are articles that by their very nature will always be that short. Otherwise, it would be prudent to add notations so that future growth of the article is managed better. --plange 15:15, 26 October 2006 (UTC)

• bump* --plange 00:24, 27 October 2006 (UTC)

I tried to add something to the effect you were suggesting. Is this reasonable? –Joke 00:41, 27 October 2006 (UTC)

Further discussion needed

The following need discussion:

1. The proposal is labeled as "Scientific" without input from the majority of the science projects: Math and physics are only one part of Wikipedia.
2. It is labeled a guideline without broader consensus from all areas of science.
3. The proposal doesn't make clear that, in Wiki, if another editor (within reason) requests a cite, the text needs to be cited. Uncited articles might not make FA or GA. As an example, WP:MEDMOS adds to Wiki guidelines, making stronger requirements than general Wiki policies for medical articles - this is not doing that, so it's redundant, and in some places as currently worded, appears to go against the grain of current Wiki guidelines.
4. "Some statements are uncontroversial and widely known among people familiar with a discipline." Wiki isn't a textbook, and isn't only for people familiar with a discipline. It's an encyclopedia, accessed by intelligent, knwledgeable people who may just need to be pointed to a source.
5. "Such facts may be taught in university courses, found in textbooks, or contained in multiple references in the research literature, most importantly in review articles." Wikipedia has to point someone who hasn't had the university courses towards where to find the information, in perhaps a textbook. Not doing so decreases Wikipedia's reliability and reputation, since unlike textbooks and journals, anyone can edit Wikipedia.
6. Ditto for this phrase: Wiki is an encyclopedia, not a graduate text. We need to point readers to the resources. "These statements are not common knowledge, but the first should be known to anyone with an undergraduate background in physics, the second to anyone knowledgeable about condensed matter physics, and the third to anyone knowledgeable about string theory."
7. This passage appears to specifically exempt Math and Physics from inline citation: Proper citation is not cumbersome, does not detract from readability, and if this were true, we could eliminate most of the inline citations from medical and biology articles, and just claim that we cite several general review articles early on, and tell the readers to go look through entire review articles. It also makes it easier for us to misinterpret or stretch facts. It is quite common for every line in a journal-published medical or biology article to contain multiple citations, and Wikipedia should be no different.

   "The verifiability criteria require that such statements be sourced so that in principle anyone can verify them. However, in many articles it is cumbersome to provide an in-line reference for every statement. In addition, such dense referencing can obscure the logical interdependence of statements. Therefore, in sections or articles that present well-known and uncontroversial information – information that is readily available in most common and obvious books on the subject – it is acceptable to cite one or two authoritative sources at the start (and possibly a more accessible source, if one is available) in such a way as to indicate that these sources can be checked to verify statements for which no in-line citation is provided."

   Your phrasing "exempt...from inline citation" implies that there is a requirement of inline citation somewhere. Please provide a citation for this assertion. I see no statement in WP:CITE requiring cites to be inline or per-sentence, only that they should be provided and formatted appropriately for the standards of the area. —David Eppstein 21:46, 26 October 2006 (UTC)

8. This also goes against common Wiki policies and procedures - if an editor (reasonably) requests a cite, it should be provided, period:

   "If another editor requests that a particular statement be sourced, it may be advisable to add an in-line reference. If the statement is easily found in the principal references for the article, this may instead be provided on the article's associated talk page, although it is worth adding it to the article if it would prevent confusion."

   But how do you then define "reasonably"? Count Iblis 17:20, 26 October 2006 (UTC)

   Well, then we get into the joys of Wikipedia—the same issue we find with how do you define a "reliable source": by consensus. If many editors agree the request isn't reasonable, the text need not be cited. But wouldn't it be much easier to just cite it, than to go through a lengthy consensus discussion, and possibly have another editor ask for a cite sometime in the future? (This could degenerate to a rehashing of the old conversation at WP:CITE :-)) Sandy (Talk) 17:30, 26 October 2006 (UTC)

   Yes, this is the indeed the same old discussion :). To give an example, let's look at a mathematical derivation like this one. I adapted the standard derivation you can find in university textbooks so that it is accesible to people with less knowledge. So, I cannot give refs to this particular derivation. And the fact that standard derivation exist should be clear. So, all could say at the start of the section is that the following is a self contained derivation and that you can find other derivations in books such as... In another discussion someone said that I could give a ref to the original derivation by Planck, but as CMummert pointed out, the original derivation was incomplete and has all sorts of problems. Count Iblis 17:53, 26 October 2006 (UTC)

   Don't want to rehash all that - you explain that if someone asks for a cite —my concern is for much broader applications than mathematical derivations. :-) We shouldn't make and label broad guidelines to address relatively uncommon issues. Sandy (Talk) 18:06, 26 October 2006 (UTC)

9. I don't know what this means: "Since Wikipedia is not a paper encyclopedia references do not need to be as concise as they are in journals."

   In scientific journals you often see citations like "ApJ" which stands for "The Astrophysical Journal". Some journals have a page length restriction for articles, and in general space in journals is expensive. In wikipedia this is not an issue. It is useful to say this explicitely, because many editors also publish in journals and are very used to using these abbreviations. Count Iblis 17:09, 26 October 2006 (UTC)

   Well, as an example, in medical articles on Wiki, citing the abbreviation doesn't trouble me, since the PMID link is included for specifics on the exact article. I'm not sure if we are agreeing or disagreeing, as I'm not that clear on this wording, but I cite abbreviated journal names in medical articles, since the PMID gives one access to the article. Sandy (Talk) 17:17, 26 October 2006 (UTC)

Sandy (Talk) 16:59, 26 October 2006 (UTC)

This document doesn't override WP:CITE, so most of your concerns are unfounded. If someone outside wishes to evaluate an article, they should use WP:CITE to do so; there is no reason that the GA or FA reviewers must pay any attention to this document. And the document links to WP:CITE and WP:V early on. The purpose of this document is to give a commonsense explanation of how most editors inside the math and physics community interpret WP:CITE. The word "Scientific" is a red herring; the title of the article is not important. CMummert 17:08, 26 October 2006 (UTC)

The title is important: other scientific areas will assume, as they should, that it applies to them, and will quote it. And, since it doesn't override WP:CITE, how is it not redundant or worse, even of little use? Yes, WP:CITE overrides this proposal, so I am not understanding it's usefulness, or why it was added to WP:CITE as a guideline? Sandy (Talk) 17:14, 26 October 2006 (UTC)

The disclaimer at the top of the document makes it clear this is the opinion of editors in math and physics; there is little chance a biologist (for example) would think it applies to the. Moreover, it doesn't need to be linked from WP:CITE and I wouldn't mind if you removed all references to this document from WP:CITE. We both agree it is not a wikipedia-wide policy. It's use is for internal evaluation of articles by the math and physics projects. CMummert 17:18, 26 October 2006 (UTC)

It was already added to WP:CITE as a guideline; I already removed it. If you are insistent that it only applies to math and physics, and aware that WP:CITE overrides this proposal, then changing the title to a math/physics project guideline would be logical, recognizing that following this proposal may preclude GA and FA status. Sandy (Talk) 17:22, 26 October 2006 (UTC)

I see how following these guidelines might preclude GA status, since the GA arbiters have come up with unreasonable criteria demanding inline citations, which, in my opinion, lead to worse articles. But how would it be a problem for FA status?  --Lambiam^Talk 18:24, 27 October 2006 (UTC)

Only by minimally following these guidelines and then refusing to add more references during GA review would "following these guidelines preclude GA status". But no matter what, refusing to do what the GA reviewers ask will preclude GA status, so it's not an issue of whether you follow these guidelines or not. The claim has been made that the "actual" FA review standards are more strict than the written guidelines for FA review at WP:FA?; that might be where the reference to FA came in. Certainly the published criteria for FA review are not as burdensome as the published ones for GA review. To be honest, after just now comparing WP:FA? and WP:WIAGA, I can't see how it is worth the trouble to nominate any article for GA. Since meeting the GA standards would qualify the article for FA status as well,you might as well aim high from the beginning. CMummert 18:34, 27 October 2006 (UTC)

Project guidelines?

Maybe I'm missing something obvious...

If this is intended to be a guideline for work done within the scope of particular WikiProject(s), wouldn't it be much easier to simply have it reside on a (probably cross-linked) subpage of the project(s) in question? As far as I know, there's ample precedent for projects creating more specific guidelines for their own fields; and other projects could endorse (or not, as the case may be) the guideline as they chose, without anyone having to worry about what exactly "Scientific" means and what areas this applies to. Kirill Lokshin 17:24, 26 October 2006 (UTC)

I think I'm missing the same thing, with the additional confusion that WP:MEDMOS amplifies and adds to Wiki guidelines, while this appears to subtract from, which renders it ... what ... redundant to possibly useless, specifically with respect to FA/GA? Sandy (Talk) 17:27, 26 October 2006 (UTC)

The criteria for the GA and FA review process are so vague that getting an article to GA or FA must be done on a per-article basis anyway. Most articles won't be FA or GA, and so the standards there aren't relevant to internal article ratings. I agree that this article doesn't need to be linked from WP:CITE, but I disagree there is a significant chance of confusion from the title, since the article itself is so clear about the intended audience. CMummert 17:31, 26 October 2006 (UTC)

Since it is clear in the text, and if it will only be used for internal Project article rating, I don't see any reason not to rename the article to reflect that. Sandy (Talk) 17:34, 26 October 2006 (UTC)

But do you have a strong reason why it should be renamed? As I said, confusion seems unlikely. In the end, I expect this document to improve, not hurt, the overall level of citation in math and physics articles (have you looked at them lately?), a goal I am sure everyone supports. CMummert 17:40, 26 October 2006 (UTC)

I'm agreeing with Kirill: if it's an internal math/physics Project guideline, it should reside internally, and not be titled in a way that other science areas may think it has more general consensus or applicability. I would be concerned to see the items I enumerated above being applied to medical or biology articles. I don't understand the resistance to simply having this be an internal Project guideline, for rating articles, minimizing the risk that other areas of science will think it applies to them? If it remains as it is now ("Science" "guideline") it needs a disupted tag; as an internal guideline, it seems perfectly acceptable. I see no problem with math/physics setting up internal guidelines for article ratings. Sandy (Talk) 17:44, 26 October 2006 (UTC)

These are intended to be consistent with WP:CITE but to provide particular guidance for science articles. Ignoring the issue of whether all science WikiProjects have been involved (they haven't yet, but that's being addressed), do you have a problem with this? Why all this talk of what trumps what? Why the opposition to people with particular knowledge of a subject writing down their ideas for how to write articles on that subject? This shoudln't be a power struggle. -- SCZenz 20:55, 26 October 2006 (UTC)

Mmm, my concern is that the guideline doesn't really make the scope of its application explicit. If it were explicitly limited to articles under, say, WikiProject Physics or WikiProject Mathematics, that would be one thing; the scope of those projects is such that there is fairly little intersection with other (particularly non-scientific) fields. But this guideline has been, at least in my understanding, designed to be adopted across all "scientific" projects.

In particular, is this guideline meant to apply to topics involving applied science or engineering? And, if so, have you considered the implications of encountering, on those articles, other projects that have an entirely different understanding of WP:CITE should be interpreted? Kirill Lokshin 21:17, 26 October 2006 (UTC)

The scope in terms of subject matter depends on the opinions of other WikiProjects, and they need to be consulted, and they will be. It was written with the concerns of physics and math writers in mind, and I expect that it may very well be modified as part of that process. I think that's an important discussion to have, and a good one. What concerns me is that some editors want to stuff these proposed guidelines back in the physics projects apparently because they want to be clear that WP:CITE trumps our ideas. If any part of WP:CITE disagrees with what's in here, that part was probably written by people who don't understand the issues involved in science writing in general or physics/math writing in particular. I don't see why anyone would be keen on making sure that those peoples' views trump the views of the subject experts; we should, rather, have a dialogue about the content of this page. -- SCZenz 21:26, 26 October 2006 (UTC)

First, SCZenz, thank you for attempting to maintain this as a proposal, rather than guideline, as that seems the most reasonable way to keep the heat down low. It moved rather quickly to guideline status, considering its limited exposure. Second, I am hesitant to rehash the entire discussion already had at WP:CITE, where many of us laid out specific examples of why this doesn't fly for all areas of science. I'll let the rest of you work on it for a while, but I wouldn't categorize myself or Kirill (for example) as people who don't understand the issues involved in science or math writing. Sandy (Talk) 21:31, 26 October 2006 (UTC)

If these don't fly for all areas of science, then we'll change the name of the page appropriately. The idea for this came from WikiProject Chemistry, so I suspect they'll be ok with them. For Biology, maybe not. For Engineering, maybe not. What I'm opposed to is making these internal WikiProject guidelines; they are not. They are guidelines written by WikiProject editors that are intended to have some weight outside the WikiProject.

Forgive me if I implied you don't have any knowledge of issues in science writing; I shouldn't have done so. However, it does appear that if you have personal knowledge of physics or math writing, it isn't reflected in your contributions. That being the case, I would hope you'd give those of us with such knowledge the benefit of the doubt in creating guidelines that are appropriate to the articles we write. If there are conflicts between this page and WP:CITE, they should be reconciled through discussion; nothing should "trump" anything. -- SCZenz 21:51, 26 October 2006 (UTC)

Disputed

Since it has now been re-added to WP:CITE, a disputed or proposal tag should be added. There is not broad consensus for this to be a scientific guideline at WP:CITE. Sandy (Talk) 17:46, 26 October 2006 (UTC)

Here is a compromise proposal. Would the following changes appease everyone:

1. Don't link to this article from WP:CITE
2. Do link to it from the per-project pages
3. Change the name to something like Wikipedia:Citation guidelines for Physics and Mathematics.

Let's slow down the discussion to give editors who have real lives a chance to respond this evening, before making any changes. CMummert 18:16, 26 October 2006 (UTC)

I'd be happy with all of that, but additionally, hope that the page will make clear that the proposed wording does override WP:CITE, and something about either that it only pertains to internal article ratings and/or could preclude GA/FA status. In fact, if it says all of that, I wouldn't care if it is linked to from WP:CITE, although linking to something from CITE that is overridden by CITE doesn't make a lot of sense to me. I'm sorry to have had to place the disputed tag, but felt it necessary since it was re-added at WP:CITE. Sandy (Talk) 18:26, 26 October 2006 (UTC)

I don't mind the steps of your compromise in principle, but I remain concerned that leaving it out of WP:CITE will allow overzealous editors to conclude that their own interpretations of WP:CITE, in which standard textbook facts and examples still require per-line citations, should override this proposal's statement that such citations can be done on a per-article basis. If that zeal prevents articles from being GA that's the loss of the GA process more than ours, but if it causes math and physics articles to be peppered with fact templates it's a problem. —David Eppstein 18:41, 26 October 2006 (UTC)

The purpose of this document is not to override WP:CITE, which wouldn't be possible anyway. The purpose is to describe some good practices to help editors see how to include citations in technical articles. As the intro says,

"Given the increasing importance that the Wikipedia community puts on footnotes and other in-line citation systems, there is a need to clarify how these guidelines should be implemented in practice."

The GA and FA review groups have their own requirements that can be dealt with on a per-article basis for those few articles that are nominated. I don't think technical articles will ever be peppered with fact tags, whether they are well sourced or not. CMummert 19:01, 26 October 2006 (UTC)

I mayhave to partially retract this statement after looking here. CMummert 01:05, 27 October 2006 (UTC)

That may not have been a technical article, but this one is. —David Eppstein 03:20, 27 October 2006 (UTC)

I'm not suggesting that it override WP:CITE itself. But the policy itself can be interpreted in different ways, as past editing disputes have made clear. What this proposal may be able to do is to provide guidelines for interpretation of the policy as it applies to math and physics, and by so doing preempt alternative interpretations of that policy, as well as to provide precedent in case editors of WP:CITE wish to make those other interpretations into a more concrete part of the policy. But if the proposal is not somehow connected to WP:CITE it would have less ability to serve those purposes. —David Eppstein 19:06, 26 October 2006 (UTC)

I have some comments on this, which I will describe at length below. I removed the disputed flag but also the added section to WP:CITE (it is still mentioned in the see also section, but I clarified that it is a proposal). If this annoys you, reinsert it, but I don't think anyone argues that this page has a Wikipedia-wide consensus so stating that its status is disputed doesn't make a lot of sense to me. –Joke 19:02, 26 October 2006 (UTC)

The page was added to WP:CITE earlier today as a "guideline"; now that it's more correctly listed as a proposal, and moved to See also on WP:CITE, the dispute tag is no longer needed. I'm still concerned about the "Science" title; math and physics are a small part of science on Wiki. More later, Sandy (Talk) 19:45, 26 October 2006 (UTC)

I don't think "small" is what you mean here. Obviously they're not anywhere close to all of science, which is sufficient to prove your point, but I do object to the word choice. -- SCZenz 03:09, 27 October 2006 (UTC)

Response to Sandy's criticisms

The whole intent of these guidelines was to move the debate away from these polarizing questions like: "every statement needs an inline reference, yes or no?" and "common/uncontroversial/undergraduate level knowledge ought not be cited", which I strongly believe are both red herrings. (Although both are answerable. The first is: "in some cases, yes." The second: "Of course it should verifiable.")

Title

My proposal for the title is to leave it at this title, with a banner clarifying its status, until we can see whether we can achieve broad consensus or not. If only the editors from the Physics and Mathematics WikiProjects are happy with these guidelines, then I would be happy moving it. I don't see how this can be controversial – plenty of proposed policies and guidelines (one relevant example is Wikipedia:Attribution) use this system, and I don't see why it should be particularly confusing in this case. It is assumed that the reader will look at the banner at the top of the page. I think this is an adequate response to Sandy's concerns 1 and 2, above.

Uncontroversial knowledge

Sandy's criticisms 4–7 seem like an obtuse reading of the guidelines to me. The point is that the density of inline citations for an article like cosmic inflation should be different from those for an article like derivative (mathematics), where most of the article can be found in standard textbooks. You say that many articles in medicine and biology have citations for every sentence. I would be surprised if that were true: I've only read a handful of articles outside of mathematics and physics, but those I have read are certainly not that densely cited. But that's not really the point. Cosmic inflation, whatever it's other flaws, is very densely cited (although not densely enough in parts). That is as it should be. Other articles, like special relativity don't need to be nearly as densely cited, because it is usually sufficient to provide one or two book references at the beginning. I really just don't see this as controversial, and it certainly doesn't seem to go against WP:V and WP:CITE.

Editors requesting a cite

(Sandy's criticisms 3 and 8.) At no point do the guidelines say that a citation should not be provided if another editor requests it. It merely suggests that editors may think it is more reasonable to provide it on the talk page if it is to standard material.

Relationship to WP:CITE and WP:V

They clearly take precedence. This is stated in the preamble. I don't think anything in the guidelines directly contradicts these. If you do, please point out the relevant sections in WP:CITE and WP:V and we can talk about it.

Relationship to WP:GA and WP:FA

I certainly hope articles can be elevated to this status when they are written in accordance with these guidelines. I don't think anything in these guidelines is incompatible with the criteria listed for the FA or GA designations. My own attitude, and I think the attitude of many other editors in the involved WikiProjects, is that if it is not possible to write an FA or a GA using these guidelines, then it would be unfortunate.

Concision

(Sandy's criticism 9.) I think the main point here is that it is standard practice in many academic journals to omit the titles of articles. This is a bad choice on Wikipedia. Also, some journals have impenetrable abbreviations for names which might cause people confusion, like JETP Lett. and MNRAS which you might want to expand: of course, if you provide links, then this might not be necessary.

Finally, let me note that these are meant to be guidelines, not policy: "it is not set in stone and should be treated with common sense and the occasional exception." –Joke 19:34, 26 October 2006 (UTC)

Regarding "editors requesting a cite": I see no problem with mentioning explicitly in these guidelines that a requested cite must be provided. It repeats the info in WP:CITE, but it's an important point and requires only a few words. For instance, we could write:

"If an editor requests that a particular statement be sourced, that request must be fulfilled. It may be advisable to add an in-line reference …"

Regarding Sandy's point 7: I'd say that providing an inline reference to every statement is cumbersome and does hurt readability. In fact, I'm having trouble seeing how you can disagree with this. It costs time finding all the necessary details and typing all the references in the proper format, doesn't it? All the blue-coloured superscript numbers are extra information that most readers aren't interested in, and hence they are distracting, aren't they? (Of course, it doesn't follow immediately from this that inline references are not required.) -- Jitse Niesen (talk) 13:31, 27 October 2006 (UTC)

This can be problematic when a layperson requests a citation for a statement which follows from a previously cited work. I think that in such cases the fact that the layperson did not understand this must be addressed. But this can be done e.g. by modifying the text, you don't necessarily have to cite the ref. again. This is not different from how scientific articles are written. If you write a big introduction where you discuss someone else's work, you don't cite every sentence. Count Iblis 13:49, 27 October 2006 (UTC)

Count Iblis, are you replying to me? We agree as far as I see, but your text suggests otherwise. -- Jitse Niesen (talk) 14:30, 27 October 2006 (UTC)

I was replying to the statement you quoted. :) Count Iblis 15:56, 27 October 2006 (UTC)

May I point out that neither WP:CITE nor WP:V (and also not the proposed WP:ATT) state that a source must be provided if requested. Such a clause will be meaningless unless it specifies who is obliged to provide that source. As a case in point, suppose an editor was to request a citation for the claim in our article Gravitational plane wave that "a(u),b(u) can be any smooth functions". Upon whom is it then incumbent to provide that citation? Jimbo Wales? Joke137? Me? The editor who requested the citation?  --Lambiam^Talk 19:55, 27 October 2006 (UTC)

Err, it's right there in the little box at the top of WP:V:

3. The obligation to provide a reliable source lies with the editors wishing to include the material, not on those seeking to remove it.

Kirill Lokshin 04:33, 28 October 2006 (UTC)

It does not state that if a source is requested it must be provided, just as the policy does not state that unsourced material must be removed. And further, can you explain how that would apply to the specific example I gave? The smoothness statement is already there. It is not that I "want to include" it. But I think an editor who removes this existing material is possibly intellectually challenged, or else not acting in good faith and trying to prove a point (or both). Then what? In every article I look at I see unsourced statements. I actually fear that we may get into senseless feuds in which valid but unsourced material is removed out of retaliation for similar removals.  --Lambiam^Talk 09:34, 28 October 2006 (UTC)

Providing a reliable source is not the same thing as providing a citation. If there are reliable sources which the reader can use to verify the information, and those sources are listed in the article, then the editor has done her job and the box at WP:V is satisfied. --ScienceApologist 04:38, 28 October 2006 (UTC)

Minor issues about eponyms

Three issues with "A related issue of attribution is eponyms...if Wikipedia has an article about the topic...then these articles must cite the original papers" - first, sometimes things are given names of people who actually did little or no work on the subject such as Pell's equation. We should make clear that in those cases a citation to the named author is not required. Second, it should make clear that if the eponym was not used by the original author (for example Euler's constant) then a citation should be given it its use by that name. Third, it must be emphasized that this requirement in no way requires Wikipedia to rely on that paper for any information about the topic (I don't want this to become a loophole for getting cranks to use out of date papers and such). Other than that it looks good to me. JoshuaZ 00:50, 27 October 2006 (UTC)

I more-or-less agree with all these points, and indeed had been thinking about the first one myself. I think it is important to cite the works of the people named in the eponyms, even if they are not principally responsible, to show where the name came from. (Two interesting examples are that Kaluza-Klein theory originated with Gunnar Nordström and the FLRW metric, for which it seems it is acceptable to use any subset of those four letters.) But we should be clear that it is even more important to cite the work of the people historically acknowledged to have made the discovery. In the second case, I think that finding the first use of a name could be very hard in some cases (not many people name things after themselves, or even use the name once it is so-named). Third point, I agree completely. I haven't read half of the papers I've cited on Wikipedia as the acknowledged historical origin of an idea. –Joke 01:00, 27 October 2006 (UTC)

DOI numbers?

One suggestion to come out of a message I left for the Chemistry WikiProject is that this guideline should suggest the use of DOI numbers [1]. Apparently they're widely-used in Chemistry; as far as I know, they are not commonly used in physics. Perhaps we could add a suggestion that citations include some numerical identifier for ease of looking the article up on the internet—this would be DOI number for Chem, but probably arXiv number for Physics. Any thoughts? -- SCZenz 14:12, 27 October 2006 (UTC)

It seems to me that those are like ISBN numbers for books and should be treated in the same way. There is a template Doi-inline to make magic links out of DOI numbers, and it should be possible to make a template for arXiv links. CMummert 14:35, 27 October 2006 (UTC)

A template for arXiv links already exists; it is called Template:Arxiv (surprise!). A possible problem is that the version in the arXiv is not always the same as the published version. -- Jitse Niesen (talk) 15:06, 27 October 2006 (UTC)

Somewhat related, for math articles do we maybe want to recommend inclusion of MR numbers? JoshuaZ 15:09, 27 October 2006 (UTC)

I think these are brilliant suggestions. I have done my best to add these in, although to be honest, I am only familiar with high energy and astrophysics, where people don't seem to use a uniform database, but rather use first the arXiv, then the Astrophysics Data System or SPIRES. Of course, it might almost be better to prefer ADS or SPIRES to the arXiv, since they provide a link to the article at the journal's servers as well, for those who have access. Anyways, what I'm saying is the suggestion I added could potentially be reworded or improved. –Joke 15:27, 27 October 2006 (UTC)

I've been adding both doi's and arxiv id's when I find them for the math articles I've been editing, using {{cite journal ... | doi = ... }} and {{cite journal ... | id = {{arxiv | archive = ... | id = ...}}...}}. Both create links to the paper though the doi links usually require a university ip address to be able to do anything useful with the link. If we had a similar MR template I'd probably use that too. But (edited to add): as far as I know, MR searches also require a university ip, so asking most editors to supply them may be unreasonable. —David Eppstein 15:35, 27 October 2006 (UTC)

The current wording calls the use of MR's "helpful" but doesn't require it. JoshuaZ 15:56, 27 October 2006 (UTC)

Of course it cannot be required, since not all articles have an MR number.  --Lambiam^Talk 20:29, 27 October 2006 (UTC)

It also is a pain to look up MR numbers, especially for the people who don't have access to mathscinet. These guidelines should encourage citation, not discourage it by "requiring" that editors do extra work. Also see my post on should versus must below. CMummert 20:49, 27 October 2006 (UTC)

It is a pain, I agree, and I doubt I'll be doing it unless I have the MR number handy. However, MR numbers can be helpful. It is of course an issue that they're useless unless you have access. I'm not sure though that the current wording can be improved; I cancelled my attempt because it got too wordy. -- Jitse Niesen (talk) 04:00, 28 October 2006 (UTC)

Interesting. I never realized you could use the arXiv template inside the cite journal template without getting into trouble. I guess I'll start using that template again. I had been using url=, which is misleading. What I would really like, although it is unlikely to ever happen, is some kind of black box I can easily put a bibtex entry into and get a citation template out of, because the bibtex entries can easily be found at SPIRES and ADS. –Joke 15:51, 27 October 2006 (UTC)

In medicine, the most widely used database is PubMed, and we use the magic word PMID (PubMed ID).--Steven Fruitsmaak (Reply) 19:16, 29 October 2006 (UTC)

must versus should

I changed the word must to should in the last edit by Joke137. In general, the word must should be avoided in guidelines. The previous version said

If an editor requests that a particular statement be sourced, that request must be fulfilled.

That is stronger than WP:V, which uses the word should, and moreover isn't true. The fact could already be sourced higher in the paragraph, in which case the request was misguided. Or the statement could be removed from the article or reworded so that it becomes covered by a source higher in the paragraph. In any case, no one editor can force any other editor to do anything; this is a voluntary system.

The real difficulty is separating unneeded citation requests (see American football) from good-faith requests for facts that require citation. CMummert 15:21, 27 October 2006 (UTC)

Fair 'nuff. –Joke 15:23, 27 October 2006 (UTC)

I have quoted an example of this at Wikipedia_talk:Citing_sources#Another_footnote; in-line footnotes are sometimes simply redundant. Septentrionalis 03:28, 29 October 2006 (UTC)

not just math and physics

I'm a biochemist among other things, and the current version of the general guidelines on the page would be non-controversial in both chemistry and in biology. The advice about what not to cite is reasonable; the advice about using full journal names--or, at the least, unambiguous self-expanding abbreviations-- is good advice, considering WP articles are not written for specialists, and will be used by the general public to obtain copies of articles. The advice about standard identifiers is particularly important, and most people here who know their subject are already following it. The beginners will need to be instructed and reminded. Biomedicine is fortunate in that most of the articles do have freely available abstracts and metadata. Those of us with access to research libraries should be prepared to help out the ones who do not, rather than blaming them for the inadequacies of the publishing system.

(The details of reference style & their placement is another matter--there is no standard. Every journal does it differently, especially in biology where they is no one single major society. Authors and those advising them on style know to just copy whatever style the proposed journal uses & not make a fuss about it. In a medium such as this, I see no particular reason for consistency except within an article or a group of related articles, but I know that's not the generally accepted view. )DGG 06:01, 30 October 2006 (UTC)

The more science areas can subscribe to these guidelines, the better. For maths and physics we had a poll on the projects' talk pages to check that we had consensus: Wikipedia talk:WikiProject Mathematics#Citation guidelines proposal and Wikipedia talk:WikiProject Physics#Citation guidelines proposal. If you think consensus can be had for chemistry and biology too, assuming they have active projects, why don't you start a similar poll on those projects' talk pages (directing discussion beyond simple comments to here)?  --Lambiam^Talk 13:08, 30 October 2006 (UTC)

While the general info may be supported, I think the examples need some work. While there is nothing specifically wrong with having only physics related issue + one chemistry, it may be confusing to editors so it would be best if examples could be from other sciences like biology as well Nil Einne 13:28, 5 November 2006 (UTC)

Absolutely. I picked many of the examples from fields I am familiar with, but it would be much preferred if they could be more evenly chosen from the disciplines. –Joke 19:43, 5 November 2006 (UTC)

Echo Sandy

Sandy's points are virtually identical to ones I would make (especially 4, 5, 6, and 7). I have two questions:

• First, who is the audience? Is it people with a graduate degree in math or physics, or is it possible that a high school student might have to write a two page essay on string theory? I'd actually expect more of the latter—people don't read encyclopedias because they already know about a topic, they read it to learn. After all, if someone has already read a textbook on special relativity, they certainly don't need to read our article on it. If, however, something is obvious to all potential readers, it doesn't need a citation, it needs to be removed. That's why academic papers don't spell out acronyms and explain basic concepts—their audience already knows all that stuff. However, an encyclopedia by definition is accessible to the masses, not just experts—we shouldn't be writing literature reviews here.
• In a general overview article such as special relativity, how hard is it to get all information in a single paragraph from one source? I've written four featured articles on Welding using textbooks and other general sources, and managed to write them so that no paragraph has more than about 2 citations (thus eliminating any readability problem). I don't see why it's that hard to put a few page numbers at the end of every paragraph so people know where to look for information.

Personal experience: someone challenged my drawing of gas metal arc welding (under the "Operation" section), saying that the welding torch is leaning the wrong way. This is basic stuff, to a welder. When I wrote the article, I put citations after every paragraph, so I know exactly where all my information comes from, and I am thus able to say "take a look at page 125 of Cary and Helzer; they say that the direction of lean can vary". If I had not done included citations everywhere, I would have had to search through all my references to find that information—three huge textbooks, one smaller book, and an online manual. And what if I'm not around the next time someone wonders? It's a no-brainer—even for the "obvious", having citations is crucial. If it's so obvious that it's not worth citing, it shouldn't be in the article in the first place. --Spangineer^ws (háblame) 00:21, 1 November 2006 (UTC)

As far as who the audience is, it depends. Wikipedia aims (or ought to aim) to make as many articles as possible accessible to educated laymen. That means most physics articles should be so accessible. Some are probably better suited to undergraduates or graduate students: I can't imagine a layman ever caring to read about conformal Newtonian gauge or the scalar-vector-tensor decomposition, for example. (By the way, academic papers do all spell out acronyms, or at least ought to.) People with graduate degrees sometimes don't know what things in their own fields are, and like ordinary people, they sometimes Google for it. If they Google for it, often Wikipedia is the top hit, so we ought to make it as good a resource for them as we can. As for whether we should be writing literature reviews or not, I don't know. Many scientists find the detailed references in Wikipedia one of its most useful attributes. Just because it is not what you use Wikipedia for does not mean that it is not a reasonable usage. (Right now an article I mainly wrote, cosmic inflation is a bit of a literature review. I hope to some day get it to FA status, but that is clearly for the long haul.)

As for special relativity, entire sections can be written from one source, or even better, from just about any source. That is the point. The guidelines never say anything about things that are obvious to all readers. There is very little that the readers of Wikipedia have in common (other than some command of the English language). The article does assert that there are uncontroversial facts, and that if an article or section is written principally using such facts, then it is usually better to point readers to a variety of references early on in an article or section, rather than cite each commonplace statement, which may not always be possible or desirable. Some sections, like those about the history and experimental evidence for special relativity, ought to be dense with citations. Those describing the elementary consequences of the theory probably won't be, and I think most science editors think that is fine. –Joke 01:27, 1 November 2006 (UTC)

I haven't taken the time to keep up here (mostly because this discussion was already had on WP:CITE, there's not much new to say, and the consensus was against the math/physics minority who wanted to lower citation requirements) but the dispute tag should be added back to the article, as it is not a guideline, it is redundant at times, it is instruction creep, it does not have broad scientific support, and WP:V is the relevant policy. Wiki is an encyclopedia, not a textbook, and citations should be provided when (reasonably) requested by any other editor. Sandy (Talk) 00:45, 1 November 2006 (UTC)

Since you haven't taken the time to keep up, perhaps you missed this, but: could you please look back at your own numbered list of objections and respond to what I feel is my reasonable request for a citation on point 7? —David Eppstein 01:31, 1 November 2006 (UTC)

This is getting really tiresome. First, there is absolutely no need for a disputed tag on a page that is labelled perfectly clearly as a proposal. Second, this is not "lowering citation requirements." Third, yes, it is redundant at times. Can you name a policy or guideline that is not redundant in places? Fourth, it is a new instruction, but the point of making it a guideline is to establish a helpful standard without the rigidity of a policy. I don't think of that as "creep." Fifth, it seems to have the support of all the science editors I have asked, with the exception of you. Look at the section above. Sixth, there was no consensus on the talk page of WP:CITE. If there was any consensus, it was against the red herring of making an exception to referencing requirements for "common knowledge." These guidelines make no such exception, if you would care to read them with a bit of an open mind.

I would appreciate it if you would respond to the content. You did that above to a certain extent, but mostly it was to knock down straw men, as if including the statement "Some statements are uncontroversial and widely known among people familiar with a discipline" means that we are suggesting than in Wikipedia these statements do not require a reference. It was to provide context. The next paragraph begins "The verifiability criteria require that such statements be sourced..." At no point to the guidelines contradict WP:V or WP:CITE and at no point do they enjoin editors not to provide a citation. –Joke 01:06, 1 November 2006 (UTC)

I think it is a matter of judgement, Joke137. An article on solar physics or climate can mention the diameter of the sun. it will have previous given a reference to more general astronomy pages where the sources are. Almost always, in the relatively unscholarly work that appropriately characterizes WP, the citations will be to a secondary source at best, usually a handbook. If its an established handbook, that's good enough for our purposes. No one cites every source in every article, and I now demonstrate it:

For Albert Einstein in section 1.1 almost nothing is sourced except the name of the student who introduced him to Kant. There is a list of references to about a dozen biographies and a number of talks & TV programs, There is no indication whatsoever of where any specific biographical statement is derived. I did not mention TV programs lightly: one of the few items cited is a PBS program, with some hard-to-verify information about his mother's view of Mileva Marić. And this seems to be considered one of the best articles in WP. Returning to the scientific part, it is stated that the term 'photon" was introduced later, by GN Lewis, and the only thing that might be a ref is a citation to the article for him. Turning to that article, it is stated again a being in a "1926 paper", but the citation to the paper is not given. In facgt, nothing in that article is sourced besides one ref. to a article on fugacity. There are 3 external refs. given:

1. a two sentence item of one of Wolfram Research's page.
2. a single page from the Chemical heritage foundation
3. a half-page sketch in a tutorial of lewis acids & bases.

Stop fooling yourselves about the relevance of the WP standards as written. According to them, I could put "fact?" at the end of almost every sentence is the einstein page aexcept the direct quotes about people's opinion of his work, and the whole Lewis article is a candidate for deletion. I did not look far; I picked 2 prominent articles from among the ones "Joke" has edited of commented on. With the exception of some of the pages in math, philosophy, and some of the taxonomy, I could do likewise with at least 90% of the science pages here. (all of these I mention are in my opinion good articles.) I challenge you to find some example of following your own rules so we can see them in context. I could not find any.

Lets pick a non-biographical article, since i mentioned "photon", There are over 57 numbered refs. & 8 others; However, every single thing in the physical constants section is unsourced, except that it is massless and has no charge. The other information, of course, is in any appropriate physics textbook--but, not one is suggested. Out that paragraph goes. I took a look closer to my own field: Gene: lots of general refs, some quotes, some general refs to databases. It's a AfD.

What I think a functional definition of verifibility does mean is that challenged statements need to be verified. DGG 07:07, 1 November 2006 (UTC)

If people accuse you of proposing guidelines that violate established policies, it is not unreasonable to refer to the written formulation of these policies in order to refute the accusation.  --Lambiam^Talk 08:43, 1 November 2006 (UTC)

If the guidelines , policies and practice are so far out of alignment, it is reasonable to fix the guideline & the policies both to have some relation to the almost universal practice. What is the pt. of guidelines & polices than almost no one seems to follow? The only use is to quote them against others. Even if those here were dictators, and had the ability to compel all the editors to rework every article to conform to their guidelines/policies, it would still be impolitic, as most of those here would be unable to follow literally. As we all depend on cooperation, we need rules that people _can_ cooperate with. One of the guidlines is common sense. Perhaps it might be a good exercise to actually examine a few articles from time to time. DGG 04:47, 2 November 2006 (UTC)

I agree that almost no article adheres to these guidelines, or to WP:CITE, but I don't see that as a bad thing. The guidelines describe how we want the articles to be, not how they are. I find your last sentence disingenuous, given that all people involved spend a lot of time writing articles. -- Jitse Niesen (talk) 05:06, 2 November 2006 (UTC)

I believe a key reason that some people feel the level of inline citation is too low is that WP:CITE is extremely vague about which uncontroversial facts actually require an inline citation. Many of the articles I run into in mathematics do have references listed at the bottom, and are quite verifiable. The point of the guideline here is to give some specific, concrete, common-sense instructions about which facts should have inline citations, unlike WP:CITE. The comon sense summary of this guideline is:

• Put one or more footnotes near the beginning of the article with a reference to some good general sources, so readers know which of the references to look in for general facts.
• Put additional inline citations for direct quotes, eponymous things, statements of opinion, and controversial or disputed facts.
• If an inline citation beyond this is requested, strongly consider adding it to the article. At least, give the person who asked for a cite information about where the fact in question can be found. The person making the request is (as always) able to edit the article if they disagree with it.

Since each of those three bullets is inherently reasonable, and pretty clear, I think many editors would make an effort to follow them if they knew it was a goal. CMummert 12:18, 2 November 2006 (UTC)

Thanks, this is helpful

I now started editing based on this guideline: I find it particularly helpful as it not only encourages inline citation, which is essential for verification but at the same time it distinguishes between such footnotes and general references. That answered my question on how one can refer to sources that are only reliable for the purpose that they are cited for (necessary in case of doubts about factual accuracy, as often occurs with controversial subjects), but not generally reliable or useful as reference for the general topic of the article.

I would suggest to add to this guideline a remark about the use of the reference section for general references about the topic of the article. Harald88 10:45, 4 November 2006 (UTC)

Concerns

I'm not entirely sure that the technique of referencing the introductory sentence with general references is a good one. Unless I've read this policy, I might not realize that those are supposed to be general references, and that makes them hard to find. It seems like a more common practice in publishing to put footnotes on section and article titles when references support most or all of the statements in that region. That might not be a good idea for the web because article titles are somewhat special, and section titles appear in the TOC. Maybe it would be better to actually explicitly create a "General references" section at the bottom of the article.

In practice, this is a good idea and it is what I've been doing. If the general references are clearly annotated, then it is usually transparent that they are general references. –Joke 21:32, 5 November 2006 (UTC)

I often have trouble when there are a bunch of general references for an article and I find a particular statement that I am wondering whether is supported by anything, or if someone just added it to the article yesterday. I guess tagging that statement or asking on the talk page is the thing to do.

It would also be helpful to add to the body of a specific-claim footnote what claims a particular reference is supporting. Especially if the footnote is in an ambiguous location and you can't tell if it encompassing an entire paragraph or just one sentence. Otherwise people will edit the article text and add in information that is not supported by any reference, but it will look like it is supposed to be because they do not rearrange the footnotes or anything. And that's otherwise difficult to detect without doing a complete reference check. -- Beland 20:00, 5 November 2006 (UTC)

I'm a little confused about what to make of these two statements. While I'm sympathetic to the points you raise, at some point referencing has to be based on either trust or verification. It is often not clear how much of the text is supported by a particular reference, but I can't imagine that it is practical to specify this in the footnotes. How would you do this? Referencing in an encyclopedia that anyone can edit is a complicated problem, and ultimately I think we have to rely on the fact that most pages have dedicated editors watching them to ensure that errors are not introduced. Referencing helps these editors, and it also allows readers to verify statements themselves, but ultimately that goes only part way to solving the problem. –Joke 21:32, 5 November 2006 (UTC)

I've partially alleviated this problem in the past by including an explanatory note in the text of the associated footnote, eg "This text is suitable as a general reference", "See ch. 3 for original derivation", etc. This makes it clearer to readers and other editors which part of the preceding text is associated with the footnote. Of course, it doesn't avoid either misleading use of sources or destruction of the integrity of the citation by edit creep, but no citation guideline will. Opabinia regalis 03:29, 21 November 2006 (UTC)

I think this is covered by the Annotations section, with the example of footnote 17. Do you think this section of the guideline is clear enough? CMummert 03:39, 21 November 2006 (UTC)

Fundamentally yes, but what I had in mind is a bit briefer and more closely tied to the text rather than just giving a series of general references. I am also thinking more of situations where there is one or a small number of general but authoritative references (Molecular Biology of the Cell or Molecular Cell Biology, in that order; Inferring Phylogenies, etc.), which may not be so clearly the case for the example given. Opabinia regalis 04:30, 21 November 2006 (UTC)

I'm also concerned about the directive to use the dominant citation style of the discipline. If faithfully implemented, that will lead to different articles having different citation styles, and the encyclopedia will look somewhat sloppy as a result. I don't really care how the actual citation is formatted, and it's fine if that's field-specific (that makes it easier to successfully look up offline references, probably). But the Cite.php system described at Wikipedia:Footnotes should be used instead of Harvard-style inline references, which are intrusive and not well suited to a hypertextual medium. The Cite.php system also makes references more machine-readable, which makes it easier to implement future universal improvements in appearance and functionality in Mediawiki or CSS. -- Beland 20:00, 5 November 2006 (UTC)

Re using the dominant citation style: this is explicitly stated in WP:CITE, so any statement of the same here is just repeating what is already WP policy. I reject the claim that Harvard style citations should not be used; again, WP:CITE explicitly allows that, and we are not attempting to override it. My own preference is to use <ref> and <references/> for explanatory footnotes, in a notes section, to have those notes refer to bibliography items in a separate bibliography section using Harvard style within the footnotes, and to use inline Harvard style for references without further annotation; see e.g. Sylvester–Gallai theorem for an example of this style. Harvard style is not uncommon in Computer Science (a version of it is used in all ACM journals, currently, I believe) but the reasons I prefer this style are:

1. It allows references to be sorted alphabetically, which makes it easier to see whether some appropriate reference is already included
2. It makes the author and publication year more visible; both are often relevant information that should in any case appear within the body of the article
3. It separates out footnotes and references, which I believe serve two different purposes and should not be confused.
4. (edited later to add): Putting the full bibliographic information within the ref markers is problematic, because doing it using the citation templates takes a lot of lines of source text and makes it hard to follow the flow of the article in the source. As a consequence, authors who put bibliographic information inside refs typically format their references manually, and typically format them inconsistently. Putting all the references together in one place in the source makes it easier to keep the formatting consistent.

—David Eppstein 20:13, 5 November 2006 (UTC)

I agree with all these comments. –Joke 21:32, 5 November 2006 (UTC)

It should in general be obvious that the documentation for the introductory section is in the main body; if comments seem to say otherwise there can be standard text. DGG 04:39, 6 November 2006 (UTC)

Question on relationship to GA/FAC

I've had this proposal on my watchlist for a while but haven't been following the discussion closely. While it seems that the active discussion has wound down somewhat, I'm not clear on where this guideline stands in relation to the GA/FAC process, as it has been mentioned above that these recommendations (which are more stylistic than anything) are somehow "weaker" than general guidelines and would exclude articles following these recommendations from GA or FAC consideration. I'm not so concerned about GA - a single nonspecialist passing or failing an article is a poor data point for assessing its quality - but I'd hate to see the more technical articles getting clobbered on FAC (or not being nominated at all) due to getting caught in the crossfire of disputes over referencing standards. In particular, I'm concerned about some people adopting an implicit standard that any article referenced according to these guidelines is inherently not sufficiently well-referenced. Have these discrepancies been resolved somewhere I don't know about? Opabinia regalis 03:37, 21 November 2006 (UTC)

I can't give complete answers, but I have been following the GA talk pages for a while and here is my opinion. This guideline would not be binding on GA reviewers; the ultimate policy is WP:V, and WP:CITE has been a guideline for a long time. And GA reviewers are always free to use their own personal criteria. So it would hypothetically be possible for an article to pedantically follow these standards but somehow not meet the GA or FA citation criteria. But I think that if an editor made a sincere effort to meet the spirit of these guidelines then it would not be difficult to pass the GA or FA review for inline citations. And since this guideline never forbids inline citations, it cannot be true that merely following these guidelies prevents GA or FA status; it may be necessary to add more citations, but that would not be a "violation" of this guideline.

In any case, the purpose of this guideline is not just for GA and FA review. The purpose is to give some guidance that will, hopefully, encourage better citation in articles regardless of whether they are good or featured. CMummert 03:59, 21 November 2006 (UTC)

Encouraging better citation by providing realistic guidelines can't be a bad thing. It sounds like merely following this guideline - ie, doing only the minimum suggested - may not be sufficient but there's no reason to stick to the minimum. Opabinia regalis 05:04, 21 November 2006 (UTC)

I agree with that assessment of the situation. CMummert 12:04, 21 November 2006 (UTC)

GA currently require inline cites. Some editors think that the GA criteria should be changed the moment this proposal becomes a guideline. There has been some discussion of the issue at WP:GA/R in the review of Homotopy groups of spheres, although thats probably not the best place for a longer discussion on the issue. Last time I looked FA did not require inline cites, but does require a good referencing scheme. Lack of inline cites are are probably the no 1 reason for failing articles as its a very easy one to check for.

There is, however, a bigger problem on accessability of mathematics articles, any sufficiently modern mathematics article, needs a lot of prerequisites which the layman will not know. Manifold failed to get FA status because of this and HGoS might fail its review over this. I've been thinking that it might be appropriate to set up some peer-review type process for maths articles to reach A-class status. --Salix alba (talk) 08:22, 21 November 2006 (UTC)

Last time I looked FA did not require inline cites, but does require a good referencing scheme. Look again: "Claims are supported with specific evidence and external citations (see verifiability and reliable sources); this involves the provision of a "References" section in which sources are set out and, where appropriate, complemented by inline citations." Sandy (Talk) 14:11, 21 November 2006 (UTC)

This isn't a requirement for inline citations. It only specifies that they be included "where appropriate". Obviously if there are no inline citations appropriate to the article, there is no requirement for inline citations.--ScienceApologist 14:36, 21 November 2006 (UTC)

And this article attempts to define "where appropriate" to exempt a narrow set of articles and editors: that is the problem. Sandy (Talk) 14:41, 21 November 2006 (UTC)

Nope, this "article" does no such thing. It explains where citations are appropriate and how citations should be handled for scientific/math-related issues. That's all. It doesn't say that there are exemptions. --ScienceApologist 14:54, 21 November 2006 (UTC)

(edit conflict; response to Sandy) When I read this guideline, it seems to say that inline citations are appropriate in many cases. The first two sections give specific advice on how and where to include appropriate inline citations for textbook facts, eponymous things, numerical data, etc. I don't see how this guideline "exempts" anyone from anything. CMummert 14:58, 21 November 2006 (UTC)

It exempts what some editors say is common knowledge, known to anyone with a undergraduate degree, found in university courses or textbooks, claiming that references are cumbersome - information found in common textbooks is easily cited, and encyclopedic articles are not limited to those with a university background in the field. This language clearly exempts certain statements from citation, and has already been used elsewhere as a reason for not citing.

"Some statements are uncontroversial and widely known among people familiar with a discipline. Such facts may be taught in university courses, found in textbooks, or contained in multiple references in the research literature (most importantly in review articles). ... These statements are not common knowledge, but the first should be known to anyone with an undergraduate background in physics, the second to anyone knowledgeable about condensed matter physics, and the third to anyone knowledgeable about string theory. ... The verifiability criteria require that such statements be sourced so that in principle anyone can verify them. However, in many articles it is cumbersome to provide an in-line reference for every statement. In addition, such dense referencing can obscure the logical interdependence of statements." Sandy (Talk) 15:07, 21 November 2006 (UTC)

Knock it off, Sandy. The paragraph you are complaining about is manifestly not an exemption for entire articles from inline references, it is a statement that explains clearly why fact-tag-bombing is not going to be tolerated on science articles that are adequately sourced. If an editor wants to add citations to articles, there is nothing in these guidelines which prevents the editor from doing so. However, if some citation-Nazi want to require editors to write articles that conform to your wildest fantasies about how every sentence in articles should include an inline citation, you're going to have to reconsider what it means to be part of a community and how consensus really works. I am getting really tired of you failing to accomodate the consensus of the editors contributing to this page. So far, you haven't shown any acknowledgement of the other side while numerous editors who oppose your authoritarianism have gone so far in extending good faith as to propose this set of guidelines that aids editors in writing better articles. You need to take a breather and be reminded that you are not an arbiter of all that is Wikiright and Wikigood. --ScienceApologist 15:16, 21 November 2006 (UTC)

Civility please, SA. Neither are you the arbiter of Wiki - for that, we have ArbCom. Sandy (Talk) 16:29, 21 November 2006 (UTC)

Civility implies respecting others and responding to what they say, rather than endlessly repeating the same strawmen. It is not just about using polite words. Exhibit some civility yourself. —David Eppstein 16:35, 21 November 2006 (UTC)

Sandy left off the end of the final paragraph in the quote:

"Therefore, in sections or articles that present well-known and uncontroversial information – information that is readily available in most common and obvious books on the subject – it is acceptable to give an inline citation for one or two authoritative sources at the start (and possibly a more accessible source, if one is available) in such a way as to indicate that these sources can be checked to verify later statements for which no in-line citation is provided."

What is claimed to be acceptable here is the inclusion of inline citations in an appropriate manner, not the omission of inline citations entirely. CMummert 16:49, 21 November 2006 (UTC)

I'd expect that Sandy's standards for the medical articles that she edits don't export overly well to, say, a mathematics article, even if they are entirely appropriate for medicine. Medicine is inherently a very empirical subject in which specific facts can individually have required extensive research to establish (or still be disputed/controversial/unknown in popular culture despite being common knowledge in the medical community) and thus require individual citations. That is not a good description of the statements presented in the guideline as examples of 'uncontroversial knowledge'. Opabinia regalis 05:02, 22 November 2006 (UTC)

Recently some bizarre excesses have tainted the GA review process, with good articles (including formerly Good Articles!) failing because some idiot counts in-line citations and says "Not enough!" Many of us would like to see that stop. Meanwhile, we want to encourage editors to produce articles that are good, regardless of whether they pass (today) as Good Articles.

Academic papers and encyclopedias have a tradition of including references, for a variety of purposes of which many apply to Wikipedia as well. Our proposal tries to guide editors in that tradition.

Peculiar to Wikipedia is the absurd pretense of some editors that large numbers of in-line citations can substitute for a careful review by qualified reviewers. Our proposal does not cater to that madness; eventually reality must win.

Wikipedia insists that articles be verifiable; we agree. Wikipedia does not insist that articles be verified; that's troubling, but not something we address with this proposal.

Questions of quality belong on an article's talk page, and it can be appropriate to cite sources there in support of an argument that does not belong in the article itself. (My view is that article citations should be there to serve readers, not referees.)

We cannot guarantee that following this proposal will get an article past the GA citation counters; not even WP:CITE seems to satisfy them. We do believe this proposal conforms with WP:CITE, and with WP:FA. Note especially that in-line citations are not required:

• Claims are supported with specific evidence and external citations (see verifiability and reliable sources); this involves the provision of a "References" section in which sources are set out and, where appropriate, complemented by inline citations. —— WP:FA

In any event, if we have done our job well an article that follows these guidelines will be a better article for doing so. --KSmrq^T 10:42, 21 November 2006 (UTC)

Looking at the guideline, the example of the aldol reaction is a good suggestion and will allow articles to pass one hurdle in the standards. So I see no conflict of that with FA or GA. There may be a conflict with the section "When not to use in-line references". Those small articles will not get FA or GA due to the citation requirements. But other than that I don't see the conflict of this guideline with FA or GA. If an article is submitted to FA or GA, the reviewer(s) may still ask for citations over and above what is called for within this guideline. But I don't see a problem with that. Am I missing something? --RelHistBuff 17:04, 21 November 2006 (UTC)

I don't think you are missing anything. By the time an article is long enough to be considered for any sort of review, it will no longer meet this description:

These articles usually describe a simple result, or a common convention or notation and are, by their nature, unlikely to ever be expanded into longer articles.

So the section on When not to use inline citations won't be applicable to articles that are considered for GA or FA review, since these articles would be too short to pass review anyway. CMummert 17:14, 21 November 2006 (UTC)

For whatever it may be worth, some of us who do humanities based FA's are similarly bothered by this "please provide a citation for the year of birth, a footnote for each sentence, a footnote for the fact that Moby Dick is a "great" novel, and a footnote for the mysterious statement that Romanticism valued the genius." Common knowledge is something that some people have never heard of, apparently, and our FA's are looking less and less like the finest encyclopedia articles in the world and more and more like undergraduate research papers. This is a bad trend, a trend that can only please a civil servant's soul. Geogre 04:01, 22 November 2006 (UTC)

Thanks, that is useful input. When I see these demands I am perplexed as to how to explain them. Is it insecurity born out of lack of subject expertise? Is it laziness, because it's easy to count but hard to comprehend? Is it imitation, the clueless following the confused?

I've reviewed many papers in my professional life, and been reviewed as well; never has an acceptance decision, nor a decision about validity, involved counting citations! As with all content, we include citations to tell readers things they want to know, or need to know, not to satisfy counters. It is foolish and obstructive to include everything related to a topic, and it is also bad practice to over-cite — even in an encyclopedia. --KSmrq^T 04:53, 22 November 2006 (UTC)

Perhaps some of these Humanities-based editors should comment at Wikipedia talk:What is a good article? about the "inline citations are required" dispute. The advocates of that requirement seem to feel it is only math and physics editors who dislike it; having some humanties editors take a position would demonstrate the extent of dislike for it. CMummert 12:04, 22 November 2006 (UTC)

Firstly, it is not correct to give the impression that everyone involved in FA or GA reviews simply "count" citations. They apply certain principles. Wiki is not your normal paper encyclopaedia. Paper encyclopaedias will never suffer the equivalent of the Siegenthaler controversy. So like it or not, different standards must apply. Perhaps it is true that as we all have very different levels of understanding on how to apply the principles, there will be "counters" among reviewers just like there are stupid "beancounters" in corporate finance departments. But, IMO, citation requests are generally made in good faith following the principles set by Jimbo. Sorry about repeating this argument, because I am certain you have heard it before, but I will keep repeating it because I do believe the principles are correct. --RelHistBuff 08:10, 22 November 2006 (UTC)

Which principles exactly? Do you agree that in articles on scientific topics inline citations may be forgone in favour of a different form of citation where – according to the usual style and custom for that field of science – inline citations are not considered appropriate, or do you feel that this is a bad habit that detracts from the encyclopedic value of Wikipedia and that the correct principles dictate that the citations must be presented in inline form in sufficient abundance, even when that form is not deemed appropriate by the scientific experts who maintain these articles?  --Lambiam^Talk 11:43, 22 November 2006 (UTC)

I think that one problem is that even the reviewers who follow the spirit of the guidelines use language like "Not enough inline citations" or "Some whole paragraphs don't have inline citations" instead of "some facts that really should be cited aren't". These sentences have very different connotations to me, and the former two sound like bean counting even if they are well meaning. The reviewers might find the science editors more reaonable if the criticisms of the articles were phrased in a more scientific way. A more general problem is that the review process doesn't ensure that the person reviewing an article has any knowledge of the subject at hand, which leads to exactly the problems that one would expect. CMummert 12:04, 22 November 2006 (UTC)

The road to understanding goes both ways. Again it is matter of assuming good faith. If one were to ask "Well, what should be cited?", you will most likely get a response. I would also like to point out that physics/math authors should not assume that if a reviewer is not an expert, that person has no right to comment and request a citation. The non-expert eyes can be most valuable for the article. Anyway, I believe these guidelines will certainly help to move things forward. --RelHistBuff 13:37, 22 November 2006 (UTC)

Thanks for the vote of confidence; we will appreciate your support in helping move things forward.

I agree that it would be unfair to depict all GA reviewers as mindless citation counters, and I hope nothing I have said will be misconstrued as implying that. I do believe a handful of counters currently skew the whole process, and do so for article after article. When confronted with such a pattern of behavior, we need not fall back on "assume good faith", but must instead seek good judgment — which they demonstrably lack.

I heartily agree that non-experts can contribute. Often before I submit a paper for formal review, I ask an intelligent non-expert friend to scan it for things that may need attention. Nor do I accept the opinions of a peer reviewer as infallible. Sometimes it is clear that they have misunderstood, which I take as a hint that I may need to explain more clearly. (Also, a paper in academia may be rejected by two journals and accepted by a third, even without substantial revision. And a flawed paper — not mine! — may be published.) But eventually, even Wikipedia must rely on expert opinion.

You have said that, for Wikipedia, "different standards must apply". I hope what you mean is that different procedures may be needed. For, I believe our goal remains: to create a body of articles at least as good as a traditional work like the Encyclopædia Britannica.

To that end, we must reject the delusion that a paragraph with many in-line citations is more accurate, fair, complete, readable, helpful, and compelling than a paragraph with few or none. We applaud the goal; we seek more effective means to achieve it. Like you, I believe these guidelines move us in the right direction. --KSmrq^T 00:38, 28 November 2006 (UTC)

I'm not sure how to explain this in a pocket guide form, but the Siegenthaller article was pretty obviously in need of citation, and the failure there had been any set of eyes at all (other than someone doing New Pages patrol a few years ago). Basically, controversial conclusions and facts need citation. Now "controversial" is difficult to nail down for those who don't know a subject. Journals have peer review (which is peer expert review), so it's easy to spot the controversial statement and the new finding. We have more trouble. However, if people reviewing FA's asked for citations only when they themselves found the statement controversial, that would still be a radical reduction in this "no citations" mess. Let's trust ourselves and honestly ask for a citation when we honestly believe that a statement is unlikely or new. Even novices will spot the crazy stuff. If they don't, then Wikipedia's philosophy is a failure, as it is built on the assumption that all can see errors. Geogre 12:53, 27 November 2006 (UTC)

annotations creep

I'm not sure if the term has been used, but in some articles in various sections of WP I have seen a trend to very long annotation s, often a long paragraph, summarizing the reference in the style of an abstract (see open access for the most extreme I know of--or can imagine). Circumstances will vary, but the annotations probably should be like disambiguation items--normally one line, sometimes two, not more.

Some editors seem to like to write long annotations, either in footnotes or inside the list of references. I find that these long annotations often contain interesting information that can be merged into the main article. I think it's better to have the information added in a suboptimal place than to not have it added at all. Later, someone else can go through the article and move information from the annotations to the article itself. I'll add a sentence the guideline about this. CMummert 11:50, 21 November 2006 (UTC)

See the Textbook reference section of Physical cosmology for an example of useful annotations. But I agree with the above commenter that the annotations in Open access seem like paragraphs that wandered off from the main article. EdJohnston 04:43, 8 December 2006 (UTC)

Consensus among math and physics editors

Discussion seems to have settled down, and I suspect there is consensus among math and physics editors (at least) for these guidelines. I have placed a notice on the math and physics project pages asking for comments. CMummert 01:54, 1 December 2006 (UTC)

My general reaction is that it's an improvement, for the purposes of mathematical and scientific articles, over the general guidelines on citations, but -- still overly detailed and overly bossy. By quite a lot, really. --Trovatore 02:06, 1 December 2006 (UTC)

I'm sure that your opinion about these guidelines will be widely respected. Could you explain which parts (if any) are particularly bossy? The tone can be edited, if it would help. The overall goal is to give some more specific guidance than WP:CITE, so the detailed nature may be hard to avoid. CMummert 02:12, 1 December 2006 (UTC)

Well, here's a passage that caught my eye, for example:

If Wikipedia has an article about the topic – as in Michelson-Morley experiment, Sunyaev-Zel'dovich effect, Green-Schwarz mechanism, Alpher-Bethe-Gamow paper and Kaluza-Klein theory – then these articles must cite the original papers, even if they were not used as sources in writing the article. (It is often the case that the name established by usage does not credit the people who actually discovered something. In this case, this should be noted, with appropriate citations.)

I see a bunch of problems with this. First, not in every case in which something is named after someone, will there be an original paper by that person. If there is, it may not be a very useful paper, or it may not be clear that it's about the topic being discussed in the contemporary sense. The bit about "the people who actually discovered something" is also seriously problematic, as it's not always clear just what the "something" is. (As an extreme example, for a while a certain editor was claiming at the transfinite number page that transfinite numbers were originally considered by certain ancient Jain mathematicians. He could give citations for it, too. It was nonsense; the Jains had apparently had the notion that there were distinct infinite quantities, rather than only one, but that's the principal thing that their writings seemed to have in common with the modern notion.) --Trovatore 03:12, 1 December 2006 (UTC)

I agree that these are guidelines, not requirements, and so they should not be phrased using mandatory language. That section was too strong, so I rephrased it to emphasize that, in the end, it is the discretion of the editors whether the original sources are included. There don't seem to be any more specific "must" or "require" phrases. CMummert 03:34, 1 December 2006 (UTC)

I did not scan this guideline carefully for "musts". Squinting so that these all go away, the current guidelines seem reasonable, and in line with my expectations. I'll echo Trovatore's concerns in a different way: I don't want this guideline to be used as a stick to prevent forward progress on reasonable and uncontroversial topics. Sometimes, one simply does not have the references one might like to have; this should not prevent the creation or continued maintenance of an article. At any rate, the current guidelines seem reasonable (unlike the original, preposterous suggestions), so many thanks to all who put this into this form. linas 22:25, 1 December 2006 (UTC)

New MathSciNet template

I already mentioned this on WT:WPM but it probably bears repeating here: I created a new template, {{MathSciNet}}. I added some text to the guidelines mentioning it and the related arxiv one. —David Eppstein 18:56, 1 December 2006 (UTC)

Endorsement From The MCB Project

I am come from the MCB project to declare our endorsement of these guidelines... what's next in the process of endorsement? --Username132 (talk) 23:02, 15 December 2006 (UTC)

I don't know of a written procedure anywhere (someone else can correct me if there is one). The process that was followed in the math project was: Post prominent notices at the talk pages for the project, and wait for discussion to die down (a few days with no posts). Repeat as necessary. If consensus seems to favor the guidelines after discussion, make another prominent announcement message saying so, and asking if there is consensus. Wait to see if anyone objects. If there are no objections in a week or so, then add the project to the list at the top of the guideline. This seemed like a fair test of consensus to me. CMummert 02:18, 16 December 2006 (UTC)

Link to journals in references

Following a discussion on my talk page, I have the following question: when citing a scientific article, should we link to the journal's web page for this article (or to an equivalent service such as JSTOR) ? As an example, see Wikipedia:Scientific citation guidelines#_note-6, which links to JSTOR. The discussion was started because of Wikipedia:External links, which says "Sites that require registration or a paid subscription should be avoided because they are of limited use to most readers.". My opinion was that we should add the link (see my talk page for the details; my main argument was that if we refer to article that is under copyright, a reader will require a paper subscription or will have to go to a library anyway, and this is no better than requiring a subscription on a web site; in addition, the web site will often provide an abstract, the list of references, or even the first page of the article, so that it is interesting even for people who do not have a subscription); if an article is interesting enough that we add it as a reference, it probably fits the exception mentioned at Wikipedia:External links: site that requires registration or a subscription should not be linked to unless: [...] It has relevant content that is of substantially higher quality than that available from any other website. What is your opinion ? Schutz 21:44, 16 December 2006 (UTC)

I think we should link to the WP article on the Journal concerned. There are many articles on scientific journals. For example, the List of scientific journals in chemistry is quite long and many of them point to articles. I'm trying to find time to write more articles and remove the redlinks. However, chemistry might be more complete than other areas. Also see Wikipedia:List of missing journals and the introduction that urges us to write more articles on journals. Articles on journals have the external links. We should not have external links in citations. We should have more articles on journals and link there. --Bduke 22:01, 16 December 2006 (UTC)

Linking to the journal's page is a good point; coming back to my example above, this would give: Michel A. Kervaire; John W. Milnor. "Groups of Homotopy Spheres: I" in Annals of Mathematics, 2nd Ser., Vol. 77, No. 3. (May, 1963), pp. 504-537 — except that you would remove the link to JStor. Schutz 22:12, 16 December 2006 (UTC)

There is a difference between links in the external links section and links within the references section pointing to online versions of articles. It makes sense that, since only users with JStor access will be able to use it, we shouldn't put a link to the JStor main page in the external links section. But when a reference already contains full information about an article, so that a reader could look it up in print, and additionally has a link to the online version, readers who cannot make use of the link can still make full use of the reference, while users who can use the link will find it helpful.

Also, the guideline says that linking to a pay site is acceptable when

"It has relevant content that is of substantially higher quality than that available from any other website."

So if there is a free version of the article online, you can and should link to that, but otherwise the pay site is trivially of higher quality since it is the only online source. CMummert 22:13, 16 December 2006 (UTC)

where to link and why

Free vs paid: The free versions on author's home pages and the like are not yet considered acceptable sources for citation, and the relationship between their quality and the quality of the published article is unknown. However, they are often the only ones available to the users. The solution is to cite both together [http: etc. Publisher's version] [http: Author's free version]. In cases where the publishers version IS free, and this will increasingly be the case, then it should be clarified: [http: Open Access Publisher's version].

As a first guideline, anything on PubMed Central is free,and good enough to cite; absolutely nothing on JStor is free; all PLOS journals are free, and any BMC title starting with BMC... The lists of open access journals on WP need major updating--see below.

I can prepare a full discussion of how to find free versions that covers all the possibilities--for the moment--and I am prepared to update it as needed (which is monthly or weekly)--this is in fact my professional specialty, so I havwe the information in any case. But where on WP should it go? Comments invited--the Open Access page is a little idiosyncratic, and therefore not the best place. I will look for suggestions here.

But it is not practical for the link to go to the article on the journals. This would be an indirect link and direct links to the actual content are preferred. For many journals such as PNAS or JACS or JAMA there would be hundreds of references. It would not answer the question of free vs paid, because although a few hundred important journals are free altogether, there are thousands for which only a few individual articles are free--and the only place to give this information is at the original reference. DGG 00:51, 17 December 2006 (UTC)

Are you suggesting that a wikilink on the journal name shouldn't be provided if a WP article exists on that journal? I believe the opposite: such links should be made, not to make it any easier to find an online copy of the article, but to make it easier to find WP articles referring to content in that particular journal and to make it possible to find out more about the journal in which the content appeared. However the purpose of such links is very different from links to copies of the article itself, which I think should always be provided if possible even when the full text is not free to all: it is free to some (whose institutions have already paid for subscriptions), accessible with only a little difficulty to many (by walking into the library of the nearest public university) and often provides useful partial information even for nonsubscribers. —David Eppstein 01:50, 17 December 2006 (UTC)

We may be misunderstanding each other. I agree with everything you say--obviously, it is desirable to give a link to the WP article about the journal. for the reasons you give, and also directly to the outside article. But can we do both in a simple and non-confusing way? (I've had a discussion in another context--eBooks--about links to theexternal source vs. links to the page about the source which would then lead to the sources. There are some people who have strong feeling about which is better. But it doesn't matter too much when its just a question of going from the link to the article, and then following the single obvious external link in the article to the manufacturer. But here, if we go just to the journal, what does the user do then? Even at a major university, with everything available, the user would still have to search again for the article. Outside, even if there were a free version available, the link to the journal would 9 times out of 10 not lead to it.

I think one of the suggestion above was to then have the link to the specific article be on the journal page. This would work for the user if there were only a few articles from that journal. It would not work if there were 100.

Outside WP, when you are reading an online article, there is a link to another article. 5 yrs ago they usually went to the journal home page, and, let me tell you as a librarian, users were very unhappy. It took us & the publishers years of work, but now they almost always go to the article, via a system called CrossRef, and users see this as one of the great things about online journals.

So what we need to do is to figure out how to link to both in an intelligible way.

As if things weren't complicated enough, there's a third possibility, for biomedicine. Every bio or med article has a PMID (PubMedID), which goes to the abstract in PubMed. The abstract has a field called LinkOut,

which goes to the publisher's version, and to the free version in PubMedCentral if there is one. This lets people get the abstract immediately, which is often good enough , at least as a start. Variations of this work in economics, math, physics, and astronomy. DGG 04:51, 17 December 2006 (UTC)

articles about journals

On the more general question of articles on scientific journals, there are 6000 or so scientific journals in Web of Science, all of which can appropriately have articles. It will be a while until we can get them done. Although we can indicate on such an article whether the journal as a whole is open access, it is not practical to say this about individual articles. I would however, be glad to help organize a project for this. The idea would be to provide for a standard set of information, and for maintaining the articles. Anyone interested in working on such as project, please tell me on my talk page. If any response at all, we can start. DGG 00:51, 17 December 2006 (UTC)

A standard set of information is already covered by the use of Template:Infobox Journal. About 280 articles link to it and I am sure some journal articles do not add the template. A project sounds a great idea. Go for it. --Bduke 01:05, 17 December 2006 (UTC)

Good articles - Space weathering

I've started a discussion on Wikipedia talk:What is a good article? about whether these guidelines now hold and whether they should be applied to the the Good Articles criteria. These been a small response so far, and one editor who wants an official okeydoke from trusted members of the relevant WikiProject. Further the Talk:Space weathering article is now a GA candidate and the question is wether these criteria should apply there. --Salix alba (talk) 11:59, 18 December 2006 (UTC)

Endorsement from Wikipedia:WikiProject Chemistry

The Chemistry Project has endorsed these proposals at Wikipedia talk:WikiProject Chemistry#Wikipedia:Scientific citation guidelines. Looks as if they now have pretty broad support. --Bduke 01:42, 20 December 2006 (UTC)

I've added Chemistry to the projects listed in the header box of the Guidelines.  --Lambiam^Talk 08:43, 20 December 2006 (UTC)