Misplaced Pages

Revision as of 16:51, 24 June 2009

Manual of Style

This page falls within the scope of the Misplaced Pages:Manual of Style, a collaborative effort focused on enhancing clarity, consistency, and cohesiveness across the Manual of Style (MoS) guidelines by addressing inconsistencies, refining language, and integrating guidance effectively.Manual of StyleWikipedia:WikiProject Manual of StyleTemplate:WikiProject Manual of StyleManual of Style This page falls under the contentious topics procedure and is given additional attention, as it closely associated to the English Misplaced Pages Manual of Style, and the article titles policy. Both areas are subjects of debate. Contributors are urged to review the awareness criteria carefully and exercise caution when editing. For information on Misplaced Pages's approach to the establishment of new policies and guidelines, refer to WP:PROPOSAL. Additionally, guidance on how to contribute to the development and revision of Misplaced Pages policies of Misplaced Pages's policy and guideline documents is available, offering valuable insights and recommendations.

This is the talk page for discussing improvements to the Citing sources page.

Put new text under old text. Click here to start a new topic. New to Misplaced Pages? Welcome! Learn to edit; get help. Assume good faith Be polite and avoid personal attacks Be welcoming to newcomers Seek dispute resolution if needed

Archives: Index, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56Auto-archiving period: 14 days

• WT:CITE

This is the talk page for discussing improvements to the Citing sources page.

Put new text under old text. Click here to start a new topic. New to Misplaced Pages? Welcome! Learn to edit; get help. Assume good faith Be polite and avoid personal attacks Be welcoming to newcomers Seek dispute resolution if needed

Archives: Index, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56Auto-archiving period: 14 days

Retrieval dates: redundant for sources with official publication dates?

This subject keeps coming up. There is an extensive discussion in the archive. Please add new comments here, not in the archive. --EnOreg (talk) 14:20, 27 April 2009 (UTC)

Metadata in citations

The utility of citations is greatly enhanced if metadata is included with them. The COinS microformat is automatically produced by some citation templates. This metadata makes the citation information accessible to bots, so that the citation can be formatted and updated with URLs, object identifiers, etc automatically, reducing editor workload. It also allows readers' browser plugins to recognise citations. Readers may use plugins such as LibX to identify an online version of a source which they can access via their library's subscription, or may use scripts such as User:Smith609/endnote.js to export citation information to a reference manager.

Since metadata increases the utility and verifiability of citations in Misplaced Pages, and can be readily provided - either manually, or via a citation template - I would propose that the manual of style note that Where appropriate, citations should include COinS-formatted metadata. This metadata is automatically produced by most citation templates, and can also be added manually..

COinS is a developing web standard for citation microformatting and is widely used already; hence it makes sense to advocate the consistent application of this format.

Would anybody have grounds to object to this proposal?

Martin (Smith609 – Talk) 16:56, 24 May 2009 (UTC)

My experience with any form of programming in Misplaced Pages that supposedly follows standards is that the people doing the programming don't necessarily read the standards, and write stuff that works most of the time, but misses some less-common cases. Also, templates have abysmal error handling, and template users have no knowledge of the special cases that the template might silently screw up.

At a minimum, I would expect any mention of any standard to point to a free official copy of the standard. I would also expect draft standards to be avoided. --Jc3s5h (talk) 17:09, 24 May 2009 (UTC)

I strongly agree with Martin that template-generated metadata is a good idea in the long run. However, I don't think it is necessary to dictate this through the guidelines. To be specific, I don't think the word "should" in Martin's proposed text is the right word. If this is a good idea, then editors will adopt it (eventually). The guideline could make an argument in favor, along the lines of "Many editors believe that ... ". Of course, then the guideline would probably also have to include an argument against. Many editors are deeply opposed to citation templates. (I am not one of these.) As long as these editors disagree, the guideline must remain more or less agnostic about the use citation templates.

(Although your proposal carefully avoids recommending citation templates, I think it is fairly clear that citation templates are the easiest way to generate metadata.)

To Jc3s5h, I would argue that it is easier to correct mistakes made with citation templates than it is to find the "screw ups" buried in handwritten citations. Bots are now routinely checking the citations written with templates, seeking out the original papers or books and correcting any sloppy errors. The mistakes in handwritten citations are still waiting for someone to notice them. ---- CharlesGillingham (talk) 10:25, 26 May 2009 (UTC)

You are advocating the use of citation templates over "scratch" cataloging?! The powers that be still haven't created a proper template for MLA use, despite requests made in the past. It is surely not easier to correct a poorly formatted template. Don't get me started on the use of the ISO dating in the templates. FWiW Bzuk (talk) 11:25, 26 May 2009 (UTC).

No, I'm saying that this guideline should remain agnostic about citation templates. Forgive me for "mentioning the war". ---- CharlesGillingham (talk) 11:29, 27 May 2009 (UTC)

To CG: my argument was that metadata generated by a template will often be invisible in the article as presented to human readers, so most editors won't see any errors. Furthermore, even if they could, most editors will have no exposure to the standards that the metadata is supposed to conform to, and so will not be able to tell the difference between good and bad metadata.

I was not making any argument about screw ups in regular bibliographic data in manual vs. template citations, but since you mention it, I argue that since there is no written manual of style (comparable to the Publication Manual of the American Psychological Association for example) that sets forth how citation templates should behave, there are no criteria for editors to decide whether a template-generated citation is right or wrong. --Jc3s5h (talk) 13:40, 27 May 2009 (UTC)

Conclusion

The only argument I can see against preferring metadata is that "if it's a good idea, editors will add it of their own accord". However, I'm not sure that this is true. Everybody agrees that good spelling, well written prose, and citations are 'good ideas', but there are thousands of articles which don't contain any of these elements. The advantage of providing a guideline is that it demonstrates consensus; with a clear consensus it is possible for bot coders to automate the (somewhat tedious) task.

With that in mind, would the following wording address the points raised above?

Ideally, citations will be accompanied by metadata where appropriate. Misplaced Pages uses the COinS microformat convention, which is automatically produced by most citation templates. In articles which do not use citation templates, metadata can be added manually according to the COinS specification.

Martin (Smith609 – Talk) 14:11, 31 May 2009 (UTC)

P.S. what is MLA use?

Since no method of embedding the COinS without a template is available, this conclusion favors templates over manual citations. If you add that conclusion to the project page, I will consider it necessary to initiate a well-publicized RfC asking whether this guideline should favor template citations over manual citations. --Jc3s5h (talk) 16:04, 31 May 2009 (UTC)

You need to read his post again. Your comment about their being "no method" directly contradicts what he just said. It can be added "by hand". ---- CharlesGillingham (talk) 13:29, 1 June 2009 (UTC)

No, it can't. I read the browsed the spec, and there is no way anyone would get that correct without some kind of program assisting the process. --Jc3s5h (talk) 14:00, 1 June 2009 (UTC)

So your argument is that (1) citation templates are bad (2) the only reasonable way to generate CoinOS is with templates (3) Therefor, Misplaced Pages should not generate CoinOS? ---- CharlesGillingham (talk) 15:35, 1 June 2009 (UTC)

On a more constructive note, would you mind posting at Template talk:Citation/core and articulate precisely what is lacking about the current Citation/Core family of template? (Beyond the relentless "date" issue.) Perhaps these problems can be corrected, instantly improving hundreds of thousands of articles. (BTW, I think CItation/core is a better place to discuss these issues than here.) ---- CharlesGillingham (talk) 15:39, 1 June 2009 (UTC)

My argument is that templates have been argued about in the talk page of this guideline forever, and neither side has prevailed; thus this guideline stays neutral on the question of whether templates should be used at all. If the guideline is changed to say the ideal citation includes data that can only be reasonably generated by a template means that the guideline would favor templates, which is a drastic change from the present position. As for what template lack, and how they could be improved, the basic problem is they are templates. The people who don't like them just don't want to type extra information such as parameter names in addition to the actual contents of the citation. --Jc3s5h (talk) 16:09, 1 June 2009 (UTC)

The point of this guideline wouldn't be to force hard-working editors who add references to spend more of their time adding parameters (we definitely don't want to scare this type of editor off, as they are probably our most valuable contributors!), but rather to allow bots to add metadata to manually-inserted references. Entering the data 'by hand' in a minimalistic fashion, without templates, would continue to be fine and encouraged, but we could make such contributions even more useful if automated software had permission to add on metadata tags - using templates or otherwise. If the only reason against encouraging the addition of metadata is that it is difficult to add, then I think we should change the guideline anyway - but perhaps with a strong clause instructing people not to bite editors who don't include metadata! If no change in editor behaviour was mandated by the guideline, would you still have any objection? Martin (Smith609 – Talk) 21:26, 1 June 2009 (UTC)

(unindent) It is against this guideline to convert an article that consistently uses manual citations to templates, or to make other wholesale changes in the citation style. So a bot that goes around and changes manual citations to templates, whether adding metadata at the same time, or not, goes against the present guideline. Furthermore, a major objection to templates is the amount of space they take up in the wikitext; if metadata is added, this will make articles harder to edit, whether the metadata is in the form of templates or in some other form. --Jc3s5h (talk) 21:35, 1 June 2009 (UTC)

Okay. Here's another couple of scenarios for your consideration: would you still object if a bot converted the wikicode of a manually-formatted reference to use a template without changing the output style (except for the associated metadata)? If so, why? Secondly, if a bot could add metadata to a manually-formatted citation using under (say) 20 characters, would this be acceptable? Martin (Smith609 – Talk) 23:01, 1 June 2009 (UTC)

It isn't just the output style that matters, it is the ease of editing the input, so changing from a manual citation to a template of the same style would still be a problem (and a bot would never be smart enought to understand a manual citation anyway). If the metadata could be put in a manual citation with fewer than 20 extra characters, that might be OK. However, if the 20 extra characters were not plain language, (suppose it looked like {{cmd|qr3$tYuNBR7=&}}) the risk is that someone would come along and update the citation, not understand the metadata, and so leave the citation inconsistent with the metadata. --Jc3s5h (talk) 23:13, 1 June 2009 (UTC)

A bot would not have to put in a citation template at all. If it was able to parse the citation, it could just as well generate the COinS & put that in. Or the metadata could be placed next to the citation in a template that only rendered the COinS. Preferably, this hypothetical COinS-generating template would use the same parameters as the cite template, so that hand-entered references could be converted to use templates if there was a consensus to do so. All of this seems consistent with the guideline. I don't know where you came up with an arbitrary 20 character maximum for metadata. No policy or guideline would seem to endorse that peculiar requirement. --Karnesky (talk) 23:41, 1 June 2009 (UTC)

I would prefer your suggestion, Karnesky, but my 20 character limit was to avoid objections that inserting COinS metadata would add too much clutter to the wikicode. I guess that the bot would have to ensure that the metadata behind a short-template continued to match the written template. So where do we go from here? We seem to have a consensus that metadata is A Good Thing - how are we to determine whether to use an inelegant-but-uncumbersome short template or a human-readable template (which I think is preferable to the direct addition of a somewhat unintelligible COinS span)? I suppose that there's nothing to stop us from creating both templates, and having editors decide which they wish to use in their individual articles. Thoughts? Martin (Smith609 – Talk) 02:53, 3 June 2009 (UTC)

This all seems ill-defined and premature to me (maybe by several decades). I don't think anyone knows how to write a bot that can parse a manual citation and create metadata. A template that looks just like a citation template but only produces invisible metadata is even worse from the clutter point of view than citation templates. If there were a desire to provide software to all users to manually enter citation data into the software, and place short non-human readable data into the text, that would be a major software development project, and it would be a long time before we were ready to update this guideline (and the definition of such a project might change beyond recognition before coding started). --Jc3s5h (talk) 03:23, 3 June 2009 (UTC)

This isn't a discussion about approving a bot. That would be premature (as no bot yet exists). If someone has interest in making such a bot, that's great. They'd need to prove the efficacy of the bot through a bot request & you are free to nay-say there. I will certainly be critical prior to an approval, but there's little reason to be critical now. Contrary to what you say, there are multiple webservices & public code that parse citations. To be sure: not all manually-entered citations are able to be parsed by these tools. But some of the tools give a confidence rating of their parsing & it is completely feasible that something generally useful could be built.

No guideline I'm aware of address what you call "clutter." I find this to be a somewhat dubious objection. The primary reason for policy to allow any citation style (w/ or w/out templates) seems to be to be inviting to new contributors. If there is consensus that long templates are a real concern, we can probably use a template similar to the cite doi template, where the key used to generate the metadata is not just some arbitrary garbage. I'd prefer that we use full, multi-parameter templates, though: easier to tweak in fixes & also to convert to actual cite templates if there is consensus for that change.

I see no reason to oppose to suggesting (and not requiring) ways for metadata to be included. We show how/why to use templates & don't require people use those. --Karnesky (talk) 04:23, 3 June 2009 (UTC)

Phrasing/placement in guideline

I see no objections of allowing metadata to be added. As with the use of templates, it appears that some want to be able to keep their own style. While they may choose to use or not use the tools available, there seems to be no reason to not inform people that the tools are available. As such, I've rephrased the section on the guideline. It should probably be "demoted" in the position on the page. Perhaps right above or below the template section? I'd personally prefer, like Martin, to see it phrased a bit stronger. But I, like Collectonian, do not yet see clear consensus on this issue. --Karnesky (talk) 17:45, 6 June 2009 (UTC)

Correct me if I'm wrong, but no-body seems to have any strong objection to including metadata. While there may not yet be consensus on the best way to achieve it, I don't see any argument against encouraging its use. Martin (Smith609 – Talk) 20:22, 6 June 2009 (UTC)

Metadata potentially takes up too much room in the wikitext, making the article hard to edit. --Jc3s5h (talk) 21:24, 6 June 2009 (UTC)

Can we have a definition of 'too much room'? Would you accept 20 characters? 50? Martin (Smith609 – Talk) 03:27, 7 June 2009 (UTC)

It is not for me to say. Open an RfC if you want an idea of what the community thinks. --Jc3s5h (talk) 03:55, 7 June 2009 (UTC)

French wikipedia's method of collating information about references

The French Misplaced Pages has a really interesting way of collating information about references. Information about a reference book can be stored in the "Reference" namespace, for example fr:Référence:La vie quotidienne des Aztèques à la veille de la conquête espagnole (Jacques Soustelle), where you can find for instance a list of editions and can choose from a drop-down a range of formatting options for the reference. These references can be categorized by topic (for example this one is in fr:Catégorie:Ouvrage sur l'Amérique précolombienne). To insert the reference into an article, it's possible to use a template (in this case fr:Modèle:Soustelle VQA) to include key bibliographic data and provide a link to the list of editions in the "Reference" namespace - see for instance the article fr:Calmecac. The template aspect of this would need a bit of amendment to fit into the English Misplaced Pages, particularly bearing in mind that the template only lists one edition (and editors on different articles may use different ones), and also in terms of formatting, but the basic idea would work fine without the template, with the appropriate link to the "Reference" namespace placed after the reference is used (in cases like Checkers speech, where the reference inside the <ref>...</ref> tags is just the minimal Template:Harvnb author, date and page and with a link down to the full reference information in the "Bibliography" section - not sure how consistently those sections are named - then the link to the Reference namespace would belong by the fuller citation in that section).

I wonder whether there is a discussion somewhere, perhaps relegated to "perennial proposals", where it has been discussed to bring a similar set-up to the English Misplaced Pages? TheGrappler (talk) 22:20, 5 June 2009 (UTC)

I have been involved in such a discussion, and the verdict was that there was too high a cost in terms of performance. However, no-body presented any evidence to back this up, so it might be worth re-proposing - if it works for the French it can't be that server-crippling. Apart from having a different name-space and fancier gizmos, this approach is similar to the one used by the {{cite doi}} template. Perhaps it would be possible to roll out a {{cite isbn}} and {{cite url}} template too, which would function in a similar fashion, and avoid users having to search for references?

Also, I've had a gander at the French system and don't entirely understand it. Could you outline a little more clearly how a user goes from finding a reference to including it in their article in the 'correct' format? Martin (Smith609 – Talk) 15:18, 6 June 2009 (UTC)

I can't say that I am dead set on implementing the fr: system here but I can see some benefits from it (principally that information on book references can be centralized, that categories can be created of references on a particular topic area, and that it would be easy (although I haven't seen it done on fr:) to systematically catalogue which users have access to certain reference books (perhaps users who have a copy can just put a note on the talk page!), which might aid source-searching. I'm not sure quite to what extent this mirrors the doi template; the fact that it can show various different editions (including possibly different editions in different languages) is definitely handy. One thing that would be really good is that metadata (including DOI or microformat terms) could clearly be added in the "Reference:" mainspace, without the need for it to be inserted into every single article manually.

As far as I can see, what people currently do if they have a reference book and want to create a full-on reference using this system, is the following (NB even for most reference books, most people don't seem to bother doing this all and just reference it in the article in the normal way). The first step is to create a page in the "Reference:" mainspace, using a template they have called fr:Template:édition which contains various pieces of information about the edition of the book. Sometimes there are more than one edition listed, like fr:Référence:Message de Frolix 8 (Philip K. Dick) (press "modifier" to see how the template works). This is an example of a work where there isn't a shorthand template to cite it. If you look at fr:Philip K. Dick#Œuvres you can see that Message de Frolix 8 is listed (it's not strictly a citation but I'm sure you can see that a citation would work a similar way). Those little links ("détail des éditions") go to the "Reference" namespace, and are created using the ref parameter of fr:Template:écrit. Some other works, like fr:Référence:Marie-Monique Robin (Escadrons de la mort), have a shortcut template to cite them (listed at the top of the "Reference" page, in this case as fr:Template:Ref-Robin-Escadrons. If you edit the Reference: page you can see the link at the top of the page is inserted by typing {{Référence|modèle=Ref-Robin-Escadrons}} ("modèle" = "template"). The linked template is fr:Template:Ref-Robin-Escadrons, which is created in order to cite the source more easily (but I have no idea what you do if there are different editions of a work in Reference: space and different people are referencing different editions, e.g. US and UK editions, in articles, and want to have shortcut templates for the two different editions; will have to find a more experienced fr: user!). You can see an example of it being used in an article here. Does that help you? TheGrappler (talk) 17:47, 6 June 2009 (UTC)

Wiki-links from footnotes

Can the standard reference templates be expanded to auto-generate wiki-links to media (and other) sources? So if for example I've never heard of the New York Times and have doubts about what this guy Blair is saying I can click over and see that it is a well respected source. Hcobb (talk) 13:36, 15 June 2009 (UTC)

I'm guessing probably not because of issues of dealing with diambigs and having to make such a feature work with the existing wiki-links most people already put in. Usually best fix is just to fix the citation to properly wikilink source names (since they are supposed to anyway). -- Collectonian (talk · contribs) 14:36, 15 June 2009 (UTC)

Good Articles

WP:GNGA#Inline_citations asserts that <ref> tags must always be placed immediately after punctuation -- a practice that I always use, because I happen to like it, but one which I thought was not actually required anywhere. In fact, I was rather under the impression that we used the original author's style/spelling/etc unless there was a good reason to change it, and that this might reasonably be construed as following the punctuation pattern.

I have long-term concerns about the Good Article process being pushed well beyond its original "pretty good" mandate into "Really, Very, Extremely, Officially Certified Good Articles" territory, and since this "essay" is recommended by the GA process and followed quite closely by both nominators and reviewers, it has a lot of power to shape "normal practice", even if that's not the stated intention.

If any of the regular editors here would like to take a look at this section, I'd appreciate it. WhatamIdoing (talk) 23:39, 17 June 2009 (UTC) who is not watching this page

should it be page or pages

I have noticed the the pywikipediabot has changed a few citation templated to say page instead of pages even when the are more than 1 page. For example ir might say pages= 22. Is this appropriate and if so where is it mentioned?--Kumioko (talk) 20:10, 19 June 2009 (UTC)

That sounds like a mistake. Notify the bot owner by leaving a message on his or her talk page. — Cheers, JackLee 03:49, 20 June 2009 (UTC)

Thanks.--Kumioko (talk) 13:28, 23 June 2009 (UTC)

citing websites that use flash

A question has arisen in the Pro Wrestling project about citing a flash based website where you can't directly link to the subpage the info is on. Is there a general take on this problem? Can you site a site when you can't directly link to the page the information is on?? MPJ-DK (talk) 12:53, 23 June 2009 (UTC)

This question has come up before – do a search of the talk page archives if you want to read the previous discussion. My view is that it would be best to find a different reference if you can, but if you really want (or need) to cite the flash website, then just provide a link to the home page and describe the subpage and how to get to it. For instance, state the title of the subpage if there is one, and explain what buttons or links a reader should click on to reach the subpage. — Cheers, JackLee 17:18, 23 June 2009 (UTC)

Citing directly after a quotation

According to WP:Citing sources#When quoting someone a citation must directly follow a quote (I'm ignoring the other location for the time being because it's irrelevant to my query.) However, does that really have to be taken so literally? For instance, look at the this section of the article that I am currently working on: Pokémon FireRed and LeafGreen#Reception. This edit is perfectly in-line with what is said on the policy page, but if you look at the article, it simply makes more sense to cite the entire thought, as opposed to attributing every sentence to the same citation, like WP:Citing sources#When quoting someone seems to advocate. It seems to me that the way the section was written did not account for the possibility that more info about the review/paper would follow the quote. Perhaps the section should be taken liberally? Artichoker 16:51, 24 June 2009 (UTC)