Template talk:Cite journal/Archive 6

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 5 Archive 6 Archive 7


Citing "nested" journal titles. Foe example Springer publish their :

  • Biomedical and Life Sciences
    • Biological Invasions

..should we simply use the most specific and abandon the top level title? Rich Farmbrough, 12:47, 31 January 2011 (UTC).

This is not a "nested journal title". Compare the PLoS or Nature journals, for example (which are technically not "nested" either anyway). That is just a sectioning/classification of the Springer websites for easier navigation. It's not part in any way of the journal title. Circéus (talk) 14:14, 31 January 2011 (UTC)

Use lowercase for "Citation type"?

The text Citation type=Journal is this template may need to be replaced with Citation type=journal because CSS classes are case sensitive. However I have not found out why this template is displaying at a larger size than Template:Cite web at Talk:Stratos 4. If anyone knows why this happened please reply here. – Allen4names 17:43, 21 February 2011 (UTC)

Never mind. I found what was causing the problem. – Allen4names 19:36, 21 February 2011 (UTC)

Bug — extra period when a quote ends with a period

When a journal cite includes a quote, the cite ends with a period (last character of the quoted text), a double quote mark, and another period. This final period (after the quote mark) should be suppressed — though only in this particular case where there is quoted text.

Here is an example (taken from United States v. Wong Kim Ark) illustrating the problem:

Woodworth, Marshall B. (1898). "Who Are Citizens of the United States? Wong Kim Ark Case". American Law Review. Review Pub. Co. 32: 556. From this refusal to permit him to land, a writ of habeas corpus was sued out in the United States District Court .... [T]hat court discharged Wong Kim Ark on the ground that he was a citizen of the United States by virtue of his birth in this country, and that the Chinese Exclusion Acts were therefore inapplicable to him. 

The same bug does not occur in {{cite court}}, for what that may be worth. Richwales (talk · contribs) 06:45, 6 January 2011 (UTC)

This is an old chestnut. Use |postscript= alone, without giving it a value: this will suppress the terminating period.
  • Woodworth, Marshall B. (1898). "Who Are Citizens of the United States? Wong Kim Ark Case". American Law Review. Review Pub. Co. 32: 556. From this refusal to permit him to land, a writ of habeas corpus was sued out in the United States District Court .... [T]hat court discharged Wong Kim Ark on the ground that he was a citizen of the United States by virtue of his birth in this country, and that the Chinese Exclusion Acts were therefore inapplicable to him. 
--Redrose64 (talk) 13:44, 6 January 2011 (UTC)
Cite court does not call upon the master cite template, so the absence of that extra period is not unexpected. — Preceding unsigned comment added by Circeus (talkcontribs) 00:20, 7 January 2011
I should have mentioned: there is a likelihood that somebody seeing the empty |postscript= may remove it, thinking that they are merely tidying up. To forestall this, it's better to use |postscript=<!-- intentionally blank --> --Redrose64 (talk) 11:41, 7 January 2011 (UTC)

Any progress on this? I'm getting the .". glitch a lot with the journal template.—Biosketch (talk) 01:28, 6 March 2011 (UTC)

AFAIK, this is not considered a bug and there is no action either in progress or planned regarding it. See info above re |postscript=<!-- intentionally blank -->. Wtmitchell (talk) (earlier Boracay Bill) 04:57, 6 March 2011 (UTC)

Can't get issue parameter to work

I hope this isn't just something really obvious, but for some reason, the issue parameter doesn't seem to display anything. For example:

{{Cite journal | title=Magazine | issue=3 | date=March 2011}} produces

"Magazine" (3). March 2011. 

Am I doing something wrong? Thanks. Torchiest talkedits 17:03, 5 March 2011 (UTC)

    • {{Cite journal|title=article | journal=Magazine | issue=3 | date=March 2011}}
    • "article". Magazine (3). March 2011. 
Nigel Ish (talk) 18:15, 5 March 2011 (UTC)
Thanks! Torchiest talkedits 07:17, 6 March 2011 (UTC)
(edit conflict) Issue understandably won't show without the journal name.
"articletitle". journalname (3). March 2011. 
---— Gadget850 (Ed) talk 18:17, 5 March 2011 (UTC)

et al. for editors

Is there a reason why et al. is not italicized for editor lists? – VisionHolder « talk » 00:15, 21 February 2011 (UTC)

Probably because it's a "loanword or phrase that has common use in English", see MOS:TEXT#Foreign terms. --Redrose64 (talk) 16:30, 21 February 2011 (UTC)
Maybe I'm missing something... but I wanted to know why it is not italicized (in both cases to which you replied). In author lists, et al. is italicized, but in editor lists it is not. For the {{Sfn}} template, it is not italicized at all. Basically, these citation templates seem to lack consistency in this matter. – VisionHolder « talk » 16:59, 21 February 2011 (UTC)
{{sfn}} doesn't italicise because {{harvnb}} doesn't - these two are supposed to be similar in rendered appearance. As for {{cite journal}}, the italicisation of et al. in the author list but not in the editor list isn't specific to this template but is a characteristic of {{citation/core}}, and so {{cite book}}, {{citation}} etc. all behave the same. --Redrose64 (talk) 21:13, 22 February 2011 (UTC)
Fixed. Ucucha 00:08, 9 March 2011 (UTC)
I'm not sure this was the right fix, since there are quite a lot of citations using |author= and et al. italicised at the end by hand. The change increased consistency in some regards but reduced it in others. Rjwilmsi 11:05, 9 March 2011 (UTC)
We'll always have some inconsistency; there may as well be some articles that use |author= with et al. not italicized. Citation/core is now at least internally consistent in its output. Ucucha 13:20, 9 March 2011 (UTC)


Med article editors love to use the Cochrane Database of Systematic Reviews, yet there's no clearly accepted way to cite their CD number. I don't see prior mention of this topic in the archives. Is it somewhere else, or just never discussed? A recent FA, Parkinson's disease used the number as |pages=CD002813, which gave decent visual presentation, but clearly is not right. That number identifies a series of reviews on one topic, often with several issues over a span of years. It is part of the DOI, suffixed by ".pub2", ".pub3" etc for successive papers. Should there not be a specific ID for this? LeadSongDog come howl! 19:35, 17 March 2011 (UTC)

How does it look with the |id= parameter? ---— Gadget850 (Ed) talk 20:14, 17 March 2011 (UTC)
  • Jansen AG, Hak E, Veenhoven RH, Damoiseaux RA, Schilder AG, Sanders EA (2009 Apr 15). "Pneumococcal conjugate vaccines for preventing otitis media". Cochrane Database Syst Rev. (2). PMID 19370566. CD001480.  Check date values in: |date= (help)
Not quite what we'd want.LeadSongDog come howl! 21:36, 17 March 2011 (UTC)
Is there a template (similar in principle to {{PMID}}) which will construct a link for the CD number? If it were named (say) {{Cochrane}}, you could put |id={{Cochrane|CD002813}} . --Redrose64 (talk) 22:24, 17 March 2011 (UTC)
As a technical note, |id={{Cochrane|002813}} would be better, as "CD" will always be present and you'd want the identifier to look like CD002813. I can easily make {{Cochrane}} if I'm given the base url for the links. I built {{Cochrane}}, which worked with all my test cases. Let me know if there are problems with it.Headbomb {talk / contribs / physics / books} 23:01, 17 March 2011 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Thank you, that's a good start. It now generates these:

  • Jansen AG, Hak E, Veenhoven RH, Damoiseaux RA, Schilder AG, Sanders EA (2009 Apr 15). "Pneumococcal conjugate vaccines for preventing otitis media". Cochrane Database Syst Rev. (2). PMID 19370566. CD001480.  Check date values in: |date= (help)
  • Dixon L, Duncan D, Johnson P; et al. (2007). Deane, Katherine, ed. "Occupational therapy for patients with Parkinson's disease". Cochrane Database Syst. Rev. (3). doi:10.1002/14651858.CD002813.pub2. PMID 17636709. CD002813. 

So the information is correct, but the presentation is still not what we'd want. Ignoring linkages, we'd seek:

  • Jansen AGSC, Hak E, Veenhoven RH, Damoiseaux RAMJ, Schilder AGM, Sanders EAM. Pneumococcal conjugate vaccines for preventing otitis media. Cochrane Database of Systematic Reviews 2009, (2):CD001480. DOI: 10.1002/14651858.CD001480.pub3.

LeadSongDog come howl! 05:23, 18 March 2011 (UTC)

I don't understand. You want to suppress PMID, capitalize DOI, move the year to behind the journal and have the identifier place as a page rather than as an identifier? Headbomb {talk / contribs / physics / books} 05:55, 18 March 2011 (UTC)
Ah, I forgot to add the PMID. I know it looks that way, but not quite. I'm following the example citations at Cochrane's site. The form is

Authors. "Title". Cochrane Database of Systematic Reviews YYYY, (Issue):ArticleNo. PMID DOI:doinumber

So I should have said:
  • Jansen AGSC, Hak E, Veenhoven RH, Damoiseaux RAMJ, Schilder AGM, Sanders EAM. "Pneumococcal conjugate vaccines for preventing otitis media". Cochrane Database of Systematic Reviews 2009, (2):CD001480. PMID 19370566 DOI: 10.1002/14651858.CD001480.pub3.

Ah, well then if that's what you want, |page={{Cochrane|001480}} should do the trick. See

  • Jansen AG, Hak E, Veenhoven RH, Damoiseaux RA, Schilder AG, Sanders EA (2009 Apr 15). "Pneumococcal conjugate vaccines for preventing otitis media". Cochrane Database Syst Rev. (2): CD001480. doi:10.1002/14651858.CD001480.pub3. PMID 19370566.  Check date values in: |date= (help)

Headbomb {talk / contribs / physics / books} 09:16, 18 March 2011 (UTC)

That presentation would be fine, but the metadata generated has page=CD001480, which is troubling, as noted above. I see that the XML display of the PubMed database record finesses this by using:
Then the human-readable presentations use that MedlinePgn visually as if it were an alias for Page. Something of that nature could work here too. LeadSongDog come howl! 13:05, 18 March 2011 (UTC)
And what's the problem with metadata giving: "page=CD001480"? Headbomb {talk / contribs / physics / books} 13:10, 18 March 2011 (UTC)
It isn't a page, it's a database record identifier for a series of reviews on one topic over many year's time. When someone uses a citation tool (Zotero or whatever) to extract that data into another work it becomes somewhat fixed in time. But the intent behind Cochrane database records is that they denote a series of publications on a topic so that a reader can always find a current expert review on that topic. Treating it as a journal page loses that idea. The relevant bit in Citing Medicine (2007) is here. Obviously we aren't following that form in all its extensive details, but it may convey the point. LeadSongDog come howl! 14:12, 18 March 2011 (UTC)
Then perhaps a {{cite cochrane}} should be made, similar to {{cite arxiv}}? Headbomb {talk / contribs / physics / books} 14:20, 18 March 2011 (UTC)
That might be necessary, though I hate the idea of proliferating such templates it might be best. The reader of the derivative work who follows the cite should (if things are done right) be able to see both the Cochrane review that (at the time) supported the derivative work's assertions and a current review to allow them to detect if that reasoning was based on an outdated basis. When standing on the shoulders of giants, it's always a good idea to pick the top giant in the pyramid on whom to stand if you don't wish to be squished. :-) LeadSongDog come howl! 14:23, 18 March 2011 (UTC)
Give me an hour or so and I'll make that template. Headbomb {talk / contribs / physics / books} 14:33, 18 March 2011 (UTC)

Alright, took a damn long while (stupid review groups with inconsistant websites and poor documentation...) but I got it.

*{{cite cochrane
 |author=Jansen AG, Hak E, Veenhoven RH, Damoiseaux RA, Schilder AG, Sanders EA
 |group=Acute Respiratory Infections
 |date=2009 Apr 15
 |title=Pneumococcal conjugate vaccines for preventing otitis media
 |issue=2 |review=CD001480 |version=3


Headbomb {talk / contribs / physics / books} 18:15, 18 March 2011 (UTC)

Wow, "you da bomb!" as they say. That looks really good. A couple of tweaks: there's an excess "10." crept into the doi from somewhere, and the ARI Group (like others) are editors, not authors. I like the idea of identifying the group, though, perhaps with ", eds." before the closing bracket? LeadSongDog come howl! 19:27, 18 March 2011 (UTC)
Fixed. Headbomb {talk / contribs / physics / books} 21:21, 18 March 2011 (UTC)


Is there some parameter to handle those? If not, one should be added. Insofar I've handled them outside the template. Preprints are often convenient when the full paper is behind some paywall, but one still wants to make the distinction and not use |url= for that. Adding |postscript= . [http://... preprint] gets repetitive. Tijfo098 (talk) 15:09, 23 March 2011 (UTC)

There is a related thread at Template talk:Citation#Preprints which is unresolved. --Redrose64 (talk) 15:18, 23 March 2011 (UTC)

PMC not creating title wikilink when url parameter present but blank, can template logic be adjusted?

If there's a cite journal with |pmc= specified this will link |title= to pubmedcentral unless there's a |url=. This is great except for a minor problem in that if |url= is specified but has no value, |title= isn't linked when it ought to be. An example was Almond. Can the template logic be adjusted, or should I sort out a bot request to perform a simple cleanup on the ~5,200 affected articles? Thanks Rjwilmsi 18:01, 28 March 2011 (UTC)


                  |{{#ifexpr:{{#time: U}} > {{#time: U | {{{Embargo|2001-10-10}}} }}
                     |{{hide in print|1=http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pmcentrez&artid={{{pmc|{{{PMC|}}}}}}}}


            |{{#ifexpr:{{#time: U}} > {{#time: U | {{{Embargo|2001-10-10}}} }}
              |{{hide in print|1=http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pmcentrez&artid={{{pmc|{{{PMC|}}}}}}}}

Fixes this. The priority for linking parameters is |archiveurl= > |url= > |pmc=. A similar change would need to be made to {{citation}} too. Headbomb {talk / contribs / physics / books} 23:35, 28 March 2011 (UTC)

Please check that has had the desired effect. — Martin (MSGJ · talk) 06:13, 29 March 2011 (UTC)
Looks good on cite journal, e.g. Amino acid ref. Thanks Rjwilmsi 06:55, 29 March 2011 (UTC)

PMID when there's a PMC

Rjwilmsi suggests that I raise a discussion point here to address my concerns about a handful of his recent edits which added a pmid to citations that already had a pmc id. An example of the issue is changing this:

to this:

For exactly the same reasons that we discourage overlinking – the dilution of high value links by low value ones – I believe that the addition of a PMID (which links to a virtually empty page at PubMed) merely clutters the reference and distracts the reader from the valuable link to the full text of the article at PMC. I was suggested to me that the PMID index, third party links from pubmed and the citation export options from pubmed are very useful, but I don't think anyone who actually views http://www.ncbi.nlm.nih.gov/pubmed/20248086 is going to find much of use there, beyond the link to http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2053251/?tool=pmcentrez. Perhaps the metadata generated by a pmid is of value to somebody, but is the effect on readers of actually displaying a very poor link outweighed by such a putative benefit? I'd be happy to hear the thoughts of others on this issue. --RexxS (talk) 00:11, 31 March 2011 (UTC)

By the argument above, why have PMC linked at all as that is clearly redundant to having the DOI available? <g> I always include both for articles such as this. My reason has more to do with search techniques than anything else. Most of us in the field only search PubMed and never do a second check of PMC for new articles as evidenced by our recent published literature reviews. (I know, bad habit and I am as guilty as anyone) The metadata available through the PMID is in many cases what I am looking for for my applications so it would save me one additional click to have it available directly. Just my 2c… --Gene Hobbs (talk) 02:10, 31 March 2011 (UTC)
That PMID links to an "OLDMEDLINE" record that clearly has not yet been cleaned up by the NLM. If the reader selects XML in the "Display Settings" dropbox, it becomes apparent that it was only added to PubMed on March 29, 2010. The balance of the metadata should eventually be linked in. But it is definitely an extreme example that should not be used to determine our normal practices. The vast majority of PubMed records are usefully linked to other articles all manner of ways, forward and backward in time. They provide consistently formatted metadata, whereas Crossref tends to reflect the various formatting the journals each use. The PMC entries provide links only to cited works, not to "cited by" works. PubMed records consistently show if the paper is withdrawn, whereas PMC ones do not. In short, both links are valuable. LeadSongDog come howl! 02:50, 31 March 2011 (UTC)
Thanks, Gene & LSD. That certainly broadens my perspective – and I had definitely missed the cited/'cited by' distinction. So, I happened to have been caught by an extreme example, and at some point in the future the page that the pmid points to is likely to be cleaned up anyway. Ok - I'll get on with putting the pmids back! --RexxS (talk) 17:48, 31 March 2011 (UTC)


This template is used to cite academic papers. Papers such as a thesis aren't published in a work, but if |journal= is not included, the title appears italicized as a longer work:

{{cite journal |last=Tang |first=Kim Y. K. |title=Automatic Generation of Board Games |url=http://citeseerx.ist.psu.edu/viewdoc/download?doi= |month=September |year=2003 |publisher=Imperial College}}
Tang, Kim Y. K. (2003). "Automatic Generation of Board Games". Imperial College.  Unknown parameter |month= ignored (help)

It should appear in quotes:

Tang, Kim Y. K. (2003). "Automatic Generation of Board Games".  . Imperial College, University of London, Department of Computing.  Unknown parameter |month= ignored (help)

Thoughts on fixing this?

The template also needs a |type= parameter so it can define the document as a Thesis, Report or whatever. It just needs this markup:


---— Gadget850 (Ed) talk 13:50, 18 April 2011 (UTC)

This template is used more specifically to cite academic papers that are published in journals. It's not appropriate for theses, monographs, papers in conference proceedings, or papers in edited volumes. If you want something general purpose that can handle all of these, use {{citation}} (but don't mix it with the {{cite journal}} series of templates as the formatting is subtly different). Alternatively, {{cite book}} would probably be a better choice for a thesis. —David Eppstein (talk) 15:51, 18 April 2011 (UTC)
OK, will just use {{cite thesis}} even though it uses a different style. ---— Gadget850 (Ed) talk 15:59, 18 April 2011 (UTC)
This is why I always just use {{citation}}. That way I don't have to remember that there is already a {{cite thesis}} and that it doesn't need to be substituted by {{cite book}}. I suppose next we'll be adding {{cite habilitationsschrift}} variants etc... Warning, though: the documentation for cite thesis says that it's really only for unpublished theses, and that for published theses you need to use a different one. —David Eppstein (talk) 16:24, 18 April 2011 (UTC)

OK, quotes can be forced by using |chapter= instead of |title=; added |type= to sandbox:

{{cite journal/sandbox |last=Tang |first=Kim Y. K. |chapter=Automatic Generation of Board Games |type=Thesis |url=http://citeseerx.ist.psu.edu/viewdoc/download?doi= |month=September |year=2003 |publisher=Imperial College}}
Lua error in Module:Citation/CS1/sandbox at line 3023: Tried to read nil global Access.

---— Gadget850 (Ed) talk 14:32, 18 April 2011 (UTC)

Names of editors appearing in wrong place and incorrectly formatted

For example:

The names of the editors should appear before the name of the journal, but after the title of the paper. Also it should probably be formatted as "Hagiya, M.; Wadler, P. (eds.):" (talk) 13:31, 8 June 2011 (UTC)

Use {{cite conference}}? --Karnesky (talk) 13:38, 8 June 2011 (UTC)
That would give
- it's the same parameters except that |journal=Eighth International Symposium on Functional and Logic Programming (FLOPS 2006) becomes |conference=Eighth International Symposium on Functional and Logic Programming (FLOPS 2006)
If that is still unsatisfactory, the positioning of editors in relation to the other items is handled by {{citation/core}}, for which {{cite journal}} is essentially a wrapper. Changing the order of display within {{citation/core}} is possible, but has a very broad impact because it affects sevaral other types of citation, such as {{cite book}}. --Redrose64 (talk) 15:19, 8 June 2011 (UTC)
That's better, but now the name of the conference is put after LNCS. It should probably be in front of it, as it is in {{cite journal}}. —Ruud 21:56, 8 June 2011 (UTC)
That would be because the "booktitle" parameter was unused, which may or may not include the conference info:
  • Sulzmann, Martin; Wazny, Jeremy; Stuckey, Peter J. (2006). "A Framework for Extended Algebraic Data Types". In Hagiya, M.; Wadler, P. Functional and Logic Programming. 8th international symposium, FLOPS 2006, Fuji-Susono, Japan, April 24-26, 2006. Lecture Notes in Computer Science. pp. 46–64. 
  • Sulzmann, Martin; Wazny, Jeremy; Stuckey, Peter J. (2006). "A Framework for Extended Algebraic Data Types". In Hagiya, M.; Wadler, P. Functional and Logic Programming: 8th international symposium, FLOPS 2006. Lecture Notes in Computer Science. pp. 46–64. 
Circéus (talk) 02:24, 9 June 2011 (UTC)
That last one is what I was looking for. Thanks. (talk) 10:29, 9 June 2011 (UTC)
Although there's an errand period between LNCS and the volume number now :/ (talk) 10:35, 9 June 2011 (UTC)
When you say "The names of the editors should appear before the name of the journal", what name do you refer to? The International Symposium on Functional and Logic Programming is not a journal, it's a conference, and Lecture Notes in Computer Science is also not a journal, it's a book series. I agree with Karnesky, this should be formatted as cite conference, not cite journal. Using citation instead of cite conference would fix the doubled-period problem, but then you need contribution= and title= instead of title= and conference=.—David Eppstein (talk) 15:48, 8 June 2011 (UTC)
Didn't know about {{cite conference}}. Would that one also be used for conference proceedings? That would fix the problem in this particular case, but I still think this could lead to problems for other citations. —Ruud 21:52, 8 June 2011 (UTC)

Request to add deadurl= parameter

Per Wikipedia:Requests for comment/Dead url parameter for citations, could the following change please be made: Sandbox Source (revision diff).

This request is one of several follow-ups on implemented change to Citation/core. The request adds an optional |deadurl= parameter that is passed down to {{Citation/core}} as |DeadURL=. Additionally, if |deadurl=no then the main |IncludedWorkURL= link sent to Citation/core will be the original link and not the archived one. This will (in Citation/core) change the archive text to lead to the |archiveurl= (see RfC or sandbox below for preview).

Here are all the test cases I could think of: User:H3llkn0wz/Sandbox3. —  HELLKNOWZ  ▎TALK 07:12, 19 June 2011 (UTC)

Yes check.svg Done — Martin (MSGJ · talk) 12:51, 20 June 2011 (UTC)

Lots of authors

How would we cope with stuff like this:


Which has about 3K authors?©Geni 20:01, 22 May 2011 (UTC)

I'd say put in as many authors as you think will do, then a "et al." It's not like ANY standard citation styke is built to accomodate more than a dozen authors or so anyway. Circéus (talk) 22:41, 22 May 2011 (UTC)
I can't believe that one paper would have that many authors - there are more names there than there are research students at most universities. Unless they wrote about one or two words each, it's got to be a mix-up. Perhaps somebody accidentally pasted in a mail group name rather than the relevant individual(s). --Redrose64 (talk) 21:34, 23 May 2011 (UTC)
No, I'm fairly sure what's going on is that more or less everyone scientifically involved wit the ATLAS experiment is listed. It's an extreme approach, though (usually large groups are given under a group name). Circéus (talk) 21:48, 23 May 2011 (UTC)
You usually would cite those as
Headbomb {talk / contribs / physics / books} 13:36, 12 July 2011 (UTC)

Title field

Make this optional as not all journal entries have titles. --Bob 21:34, 24 October 2006 (UTC)

Could you give some examples (example template call and expected output)? On which wikipedia article(s) do you intend to use that? --Ligulem 21:41, 24 October 2006 (UTC)
AIDS reappraisal --Bob 23:09, 24 October 2006 (UTC) at the moment it gives {{title}} in the output field for the Koch articles. --Bob 23:09, 24 October 2006 (UTC)
I would have thought that a title is obligatory in giving a reference and to make this optional would be a poor choice - the fact that it is displaying "{{{title}}}" I think is a good pointer to there being missing essential information. In the case given, does no one have access to the articles ? I suppose a neater alternative would be to allow an undefined title parameter to show "<? title>" (which I prefer to "?" or "???"). David Ruben Talk 23:43, 24 October 2006 (UTC)

In AIDS reappraisal I found

{{cite journal | author = Koch R. | journal = J. Hyg. Inf.| volume = 14 | issue = | pages = 319-333 | year = 1896 }}

which gives

Koch R. (1896). J. Hyg. Inf. 14: 319–333.  Missing or empty |title= (help)

Is the missing title there due to the fact the editors just don't know the title or does that source really not have a title? General question: are there really articles in journals that don't have a title? I'm not an expert in journal articles but I do have problems to imagine this. Since David sounds rather against this change request I've removed the edit request until we know more about this. --Ligulem 11:24, 25 October 2006 (UTC)

Yes, there are. In most instances they would be letters pertaining to another article. It is quite a common occurence and one that would be exceedingly easy to implement in this template. --Bob 19:38, 17 November 2006 (UTC)
Ok. I'm willing to implement that. Could you provide the new template code in a sandbox please? For example at User:Grcampbell/Cite journal? We could then do some test transclusions first (for example at User talk:Grcampbell/Cite journal), and if everything is ok, put it live. --Ligulem 19:50, 17 November 2006 (UTC)
Ok if reference really has no title, then perhaps should at least indicate what the citation is to. "Smith A (2006) BMJ 303;12:12" just really seems obscur.
  • I would prefer to see something along lines of "Smith A (2006) <Letter> BMJ 303;12:12". It is quite common to enclose the title of a paper in single square brackets to indicate that this is the English translation of the original paper's title. So use of "< >" or "( )" seems one possible way to indicate that the Title parameter is being assign something with some special meaning. Similarly one might have "<Editorial>".
  • Alternatively, and a little more substantial, would be to describe what the letter is about (usually journals will have a heading under which several letters are collected together to comment on a recently published paper). Perhaps we could have a 'Description' parameter, which in above example might be set to "Letter on antibiotic resistance" and produces this style of result: "Smith A (2006) BMJ 303;12:12 - Letter on Antibiotic resistance". it is then clear that there is no offical title, but also clear what type of reference this is to. David Ruben Talk 20:50, 17 November 2006 (UTC)
I recently cited a letter from an invited-letters column in a journal. The letter was posted online, but didn't have its own URL: there was just one URL for the whole column. FYI, I cited it using two separate citation templates combined inside one <ref> entry, like so:
<ref name="contagiousness">{{citation
|contribution=Contagiousness of the common cold
|author=Gwaltney JM Jr, Halstead SB
|separator=.}}. Invited letter in {{cite journal
|title=Questions and answers
|journal=Journal of the American Medical Association
|date=16 July 1997
|accessdate=16 September 2011}}</ref>
This generated the formatted output:
Gwaltney JM Jr, Halstead SB, "Contagiousness of the common cold",   Unknown parameter |separator= ignored (help); Missing or empty |title= (help). Invited letter in "Questions and answers". Journal of the American Medical Association. 278 (3): 256–257. 16 July 1997. Retrieved 16 September 2011. 
Cheers, --Unforgettableid (talk) 10:55, 18 September 2011 (UTC)
None of the parameters in the second {{citation}} occur in the second, so put it all in one:
<ref name="contagiousness">{{citation
|contribution=Contagiousness of the common cold
|author=Gwaltney JM Jr, Halstead SB
|title=Questions and answers
|journal=Journal of the American Medical Association
|date=16 July 1997
|accessdate=16 September 2011}}</ref>
This generates the formatted output:
Gwaltney JM Jr, Halstead SB (16 July 1997), "Questions and answers", Journal of the American Medical Association, 278 (3): 256–257, retrieved 16 September 2011  Unknown parameter |separator= ignored (help); |contribution= ignored (help)
--Redrose64 (talk) 16:58, 18 September 2011 (UTC)

ID numbers

What value is the PMC ID to the reader? Very little. About as useful as displaying the bare URL, which we don't do. The PMID (which will be present if a PMC is present) is arguably a more useful code and will always offer the reader a link to the PMC article if available. I propose:

  • If the url is missing but the pmc is present, use the pmc url for the title hyperlink and just hide the whole PMC XXX field from the user.
  • If the url is present (to indicate that the publisher's own text is freely available online) then display the PMC field as Free text available at PMC or similar wording. This would be much more obvious to casual readers that this is a very useful link rather than just yet another link to a database of abstracts or pages demanding stupid amounds of money for three sheets of A4.

-- Colin°Talk 09:44, 16 September 2011 (UTC)

While I'm here, can I also request we drop the PMC/PMID/DOI wikilinks. I know it has been suggested before. Arguments were that these are obscure codes. Well the reader will discover what those fields are for by clicking on the external hyperlink and anyone truly interested in what the code means can easily type "PMID" into their Wikipedia search box. It is horrenous overlinking and it also bloats the rendered HTML. It creates too much blue so the reader doesn't really know whether to click on the PMC part or the XXX part. -- Colin°Talk 09:44, 16 September 2011 (UTC)

Can I add my support to Colin's comments about the PMC/PMID/DOI wikilinks? Against my inclinations, I have recently had to recommend to an editor that he drop the use of citation templates in a long article to allow it to be edited. The templates were responsible for fully 70% of the rendering time for the page on the server, and this is one of those very real exceptions to the general injunction not to worry about performance. Dropping the PMC/PMID/DOI wikilinks will only improve this marginally, but in these sort of cases lots of small savings add up to performance increases. It's also worth noting that three out of the four browsers that I use regularly (Firefox, Chrome and Opera) allow me to highlight a word and Google search for it with a couple of mouse clicks. That search on 'DOI', 'PMID' or 'PMC' lead to Google results containing the relevant Wikipedia article in the top three for each case, so many readers won't even have to type into the WP search box to rapidly reach the information if they really want to know exactly what DOI, PMID or PMC means. --RexxS (talk) 17:45, 16 September 2011 (UTC)
I'm also not sure that the PMC/PMID/DOI wikilinks are very valuable, so I'd be in favour of dropping them and making the text and value link together like PMID 8759 does. I did think the auto-linking of PMC to URL was a good feature, even if it was strictly a duplication. However I disagree about not displaying the PMC number, if you want to copy the IDs for later use you need them shown and linked, not just linked. Rjwilmsi 18:57, 16 September 2011 (UTC)
Unlinking identifiers is a bad idea. It would create an extremely jarring presentation of something like
As as a create mismatches with "manual" citations using {{Bibcode}}/{{PMID}}/etc...
Either link 'em all like Bibcode: 2005JIMPT..22..132S (which is the best thing as far as readability is concerned) or link 'em all like Bibcode: 2005JIMPT..22..132S (which loses explanatory power, especially with things like JFM 0123456789 which one cannot assume that the reader knows what JFM stands for Jahrbuch über die Fortschritte der Mathematik)). Hell, I never knew what exactly a doi was until I clicked on those links. Granted it was several years ago, but the point remains that the former is more friendly to the curious non-expert and a majority of readers. People are free to tweak the behavior of this in their skins if they would rather see it differently, but we have to think of this in terms of WP:ACCESS. Headbomb {talk / contribs / physics / books} 01:08, 17 September 2011 (UTC)
I think this is the clearest of all. It removes the sea of blue problem. You can distinguish the separate links. It complies with the MOS guidelines on linking. The only reason that every single "doi" and "PMID" is a wikilink is that this template is too dumb to only link the first time. Anybody handwriting their citations would only link the first, if they felt the link was necessary. How about having some header text for references that use the cite journal templates. Something like: "For an explanation of citation codes, see the citation glossary." Editors could place this at the start of their ref section. There are WP:ACCESS problems with the overlinked version. Citations are obscure terse things and it isn't the purpose of each an every one of the hundreds of citations in a big article to teach readers how to read them. For example, we don't explain which numbers are the issue, volume or page and many editors use standard abbreviations for journal names. A citation glossary for the cite journal format could explain to readers what the components of a citation are.
And can we have a loud klaxon go off whenever someone mentions tweaking skins. Only a tiny percentage of readers are editors with accounts and only a tiny percentage of editors could contemplate tweaking their skins. It is not a solution to anything. Colin°Talk 08:36, 17 September 2011 (UTC)

I removed the {{edit protected}} for now; clearly, this isn't a simple request to change 'x' to 'y', and not 'fair' as an edit req. It's being discussed; that's fine, and I'm sure many participants in said discussion are quite able to effect any change needed  Chzz  ►  02:08, 17 September 2011 (UTC)

I restored it. The request is to revert to the old behaviour of having the PMC automatically link the title and to match the {{citation}} behavior, which has consensus. AKA undo this. The rest of the discussion concerns the appearances of certain identifiers. Headbomb {talk / contribs / physics / books} 02:45, 17 September 2011 (UTC)
OK, fair enough.  Chzz  ►  18:33, 17 September 2011 (UTC)
This may sound silly, but I like the aesthetic of doi's on display, just as I like seeing ISBNs and publisher info on display with book citations. Maybe it's mostly bells and whistles, but I think these parameters can be a sign to readers that someone has done their research. I also think they are (typically) a sign of a higher quality source. To me, if I can pick from finding the same information in a source with or without a doi, I pick the one with the doi because it looks more professional. I think displaying these parameters can be a small incentive to editors to find high quality sources. I also like for each parameter to be linked; clicking on doi was also how I learned what it was. I don't really think there is a "sea of blue" problem. Jesanj (talk) 04:23, 21 September 2011 (UTC)

Recent edit to remove URL generation from PMC parameter

Until a recent edit by Ucucha the {{cite journal}} source would automatically create a |url= parameter from the |pmc= parameter if no other URL was given. It's my understanding that this was a long-standing feature and was desirable, and done for PMC as it is the only parameter that will always link to free-access full source. Ucucha changed the logic apparently to "remove redundancy". I do no think this was an appropriate change, and as no change has been made to {{citation}} this and cite journal are now in slight inconsistency. Ucucha responded to my question that the change was not discussed. An example below:

  • citation: Selvaraj, Vimal; Plane, Jennifer M.; Williams, Ambrose J.; Deng, Wenbin (2010), "Switching cell fate: the remarkable rise of induced pluripotent stem cells and lineage reprogramming technologies", Trends in Biotechnology, 28 (4): 214–23, doi:10.1016/j.tibtech.2010.01.002, PMC 2843790free to read, PMID 20149468  Unknown parameter |month= ignored (help)

Do other editors think that the most recent edit was appropriate? My opinion is no, and that we should revert to the previous behaviour (on a techincal note it may be better to do the PMC->URL generation within citation/core). Otherwise I think we'd need to change {{citation}} too. Thanks Rjwilmsi 14:40, 12 June 2011 (UTC)

I'm sorry for not discussing the change and for not synchronizing with {{citation}}. However, I still think the change is appropriate: we simply shouldn't be putting the same link in twice. If the concern is that the fact that the PMC link leads to a free full text should be clearly indicated, we can make it say "PCM (free full text)" or something similar. Ucucha 14:43, 12 June 2011 (UTC)

The rationale that this double linking happens because this is "the only parameter that will always link to free-access full source" seems very strange to me. We similarly double-link jstor, which is generally not free-access. And we do not double-link arxiv, which is free-access. —David Eppstein (talk) 15:35, 12 June 2011 (UTC)

Wasn't the "JSTOR double linking" removed over a month ago? Circéus (talk) 19:59, 12 June 2011 (UTC)
From this template, apparently yes, but whoever did it seems to have created the same inconsistency that Ucucha is complaining about: it still double-links in {{citation}}. —David Eppstein (talk) 04:38, 13 June 2011 (UTC)
I think that'll also be my fault. Whatever the result of this discussion, it'll have to be applied consistently to both templates (and preferably, in Citation/core). Ucucha 21:37, 13 June 2011 (UTC)
Yes, I agree. (Accidentally) creating greater inconsistency between the templates is a step in the wrong direction. —David Eppstein (talk) 04:47, 14 June 2011 (UTC)
Yeah, the functionality (PMC linking) should be restored. Headbomb {talk / contribs / physics / books} 16:32, 6 September 2011 (UTC)
I disagree. I think that having the same link in both the id section of a citation and on the title violates WP:OVERLINK, so I would prefer that the only links that go on the title are the ones directly from the url parameter. —David Eppstein (talk) 18:11, 6 September 2011 (UTC)
Overlink concerns prose and wikilinks. And it also creates a discrepancy with {{citation}}.Headbomb {talk / contribs / physics / books} 18:24, 6 September 2011 (UTC)
Yes, and the point of any citation template is to aid verification. Putting the full text link on the title is consistent with other citation templates, and is an aid to readers. Jesanj (talk) 01:02, 7 September 2011 (UTC)
To those who are familiar with pmc and the other archives, the duplicate link on the title will be redundant. However, I think that from the point of view of a casual reader (who may have no idea of the difference between pmc and pm, for example), it is desirable to have a link to free, full text on the title, whenever it is available. This is a usability issue and making things easier for a significant section of our readers is an improvement to the encyclopedia. I'd therefore recommend restoring the previous behaviour, preferable at the level of citation/core. --RexxS (talk) 14:18, 7 September 2011 (UTC)
I don't mind the duplicate disappearing, but that's because I already know that what I want can be had by clicking on the PMC link. It is not generally appropriate to assume that normal readers know what I do about our citations, and RexxS has a good point about the duplicate likely being valuable to casual readers. WhatamIdoing (talk) 04:58, 8 September 2011 (UTC)
Please undo the edit. It removed links to the free full texts of sources from their titles. There is sufficient opposition to this edit, in my opinion. Thanks. Jesanj (talk) 21:42, 15 September 2011 (UTC)
Yes check.svg Done — Martin (MSGJ · talk) 14:40, 20 September 2011 (UTC)
Can we restore this functionality again? I don't understand this edit summary. Jesanj (talk) 01:21, 20 January 2012 (UTC)
I undid my last edit. I am going to remove |Embargo={{{pmc-embargo-date|1010-10-10}}} since |Embargo= is no longer supported by {{citation/core}}. ---— Gadget850 (Ed) talk 11:43, 20 January 2012 (UTC)

why is "doi" wikilinked?

this is a term that gets repeated about a jazillion times in a list of refs if someone is using them. And the blue ends up being right next to an important external link for the actual doi of the article. We are really starting to move towards having every portion of the citations linked and it is low value and distracting. The link for the article itself fine. But on top of that authors, journals, and REPEATED use terms like doi, ISBN, etc. Too much blue!TCO (talk) 05:51, 15 June 2011 (UTC)

Because many people don't know what a doi is. Likewise for all the other identifiers, like JSTOR, bibcode, MR, JFM, PMID, etc... Headbomb {talk / contribs / physics / books} 18:32, 15 June 2011 (UTC)
don't think the blue next to blue is worth it. Rather have them look it up once, than have every single ref in creation have these junk links. Really low value, especially compared to the actual article link. MOS advises against blue next to blue.TCO (talk) 18:36, 15 June 2011 (UTC)
Good for you that you know what "doi" and ISBN refers to. Did you know that JFM stood for Jahrbuch über die Fortschritte der Mathematik? Or that OCLC refers to the Online Computer Library Center? Or that OL is short for Open Library? Or that Zbl stands for Zentralblatt MATH? Or that OSTI refers to the Office of Scientific and Technical Information? Or that the SSRN refers to Social Science Research Network?
Remember Wikipedia is for everyone. What you know, someone else might not know. If the links annoy you that much, you can always create (or get someone else to create) some filters to remove them via you monobook.js / /vector.js skins. Headbomb {talk / contribs / physics / books} 19:04, 15 June 2011 (UTC)

No I don't. I don't know what doi is and I sorta hate it. but I would break down and look it up manually as needed. Not want to ruin all the legitimate, important links to offsite content, by having a junklink blue-next-to-blue for every citation in the world. Recalls to me the date wikilinking. Massive cruft.TCO (talk) 21:50, 15 June 2011 (UTC)

I think the uniform blue of the MR/DOI/JSTOR/PMID/whatever section at the end of each reference is less distracting than alternating blue and non-blue would be, actually. —David Eppstein (talk) 08:50, 16 June 2011 (UTC)
that makes no sense. You have two separate links (the junk wikilink and the external wikilink) back to back. The directions on linking specifically say not to do this as it is hard to resolve which link actions what (if they are together or separate). Also, we are starting to get more and more blue in the references. whole thing will be a mash of blue soon, impossible to tell where the real "value" links are. We have people linking journals and publishers now. What a WASTE! Bad wiki, bad, bad, bad! No biscuit!
Related to this: we certainly have an inconsistency where ISBNs are concerned. If I put ISBN 0123456789 as plain text, it comes out as ISBN 0123456789 with one blue link; but if I use a citation template with the |isbn=0123456789 parameter it comes out as Doe, John. A Book. ISBN 0123456789.  - there are two different blue links. --Redrose64 (talk) 16:19, 16 June 2011 (UTC)
I usually have interesting points mixed in with all my trolling silliness.  :) Glad you found this ISBN thing. I much prefer clicking on the first (entire) one, than having that blue next to blue, which may drive me to wrong spot.TCO (talk) 16:35, 16 June 2011 (UTC)
I agree the dual links are confusing. I often click the wrong one by mistake. Kaldari (talk) 20:01, 11 August 2011 (UTC)
Wikilinking all these abbreviations is wrong per WP:MOSLINK. It also bloats the HTML. And wikilinking the journal in every ref is also wrong. A reader may not know what a DOI is but they are perfectly capable of cutting and pasting "DOI" into the search box. We don't keed to wikilink all ten billion DOIs all over WP. Colin°Talk 13:48, 7 September 2011 (UTC)
There absolutely nothing in WP:MOSLINK that says this practice is "wrong", and if it were saying that, the problem would be with WP:MOSLINK and not the practice. Wikilinking the journal in every reference is highly desirable. If you click on [28] in Quark, you don't care that [6] and [7] references were linking to Physical Review Letters, you're concerned about [28], not [6] and [7]. Headbomb {talk / contribs / physics / books} 19:08, 17 September 2011 (UTC)
I'm still rather torn over the whole thing. While overlinking is an issue in large, extensively cited articles, it's hardly a general problem in the majority of articles. Making it even marginally more convenient for users to check a source is a very good thing. In a long article, it can be very inconvenient to have to use the search box, because scrolling up to it can make one lose one's place in the article. In contrast, with a link, one is just a ctrl-click from seeing the target in a new tab. Yes, this is just a UI/skin problem, but it is real.
On a related note, the above lead me to "http://dx.doi.org/" site:en.wikipedia.org/wiki/ an find that there's a mission there for some bot (or maybe AWB) to address. LeadSongDog come howl! 14:30, 8 September 2011 (UTC)

I think these links should be removed. They are not part of the core information, and people can always use the search box if they want to know what one of these abbreviations means. I find such excessively repeated links extremely annoying, not just these but also the link to some inane page about archiving that I keep clicking when I want to read talk page archives, the link to Help:Edit summary which for some reason gets clicked quite a lot when I try to save a page, and the link to WP:AfD from some AfD messages, that is fat and therefore more visible than the link to the actual AfD. This is a usability issue for me already. It must be much worse for users of text browsers and screen readers, who have to circle through the links when they want to click one. Hans Adler 19:24, 17 September 2011 (UTC)

Redundant to cite both issn and oclc?

The only example shown that cites an issn is the second citation under "Other examples" under Template:Cite_journal#Examples. This example cites both the issn and the oclc. But my experience is that when I cite both the issn and the oclc, clicking on the resultant links yields the identical same WorldCat page. For example:

Therefore, wouldn't it be better to cite either the issn or the oclc but not both. Wideangle (talk) 06:46, 1 August 2011 (UTC)

Even better, use neither. ISSN/OCLC are periodical identifiers, not article identifiers. Headbomb {talk / contribs / physics / books} 20:27, 11 August 2011 (UTC)
ISSN and OCLC are used for more than just periodicals and are useful for works without an ISBN. When an ISSN is used for a journal, it applies to the journal as a whole and not a single issue, so it is generally redundant unless referring to the the journal series. In the given example, the ISSN and OCLC both refer to the journal series. refers to a paper and is perfectly valid in this use. The question is— does the citation need four identifiers? It doesn't really hurt and it isn't that much clutter. ---— Gadget850 (Ed) talk 20:52, 11 August 2011 (UTC)
No, the ISSN refers to Plant Molecular Biology (ISSN 1573-5028), the journal. The article is "Transcription Profile Analyses Identify Genes and Pathways Central to Root Cap Functions in Maize" (doi:10.1007/s11103-005-4209-4). In this case, the ideal citation would be
  • Jiang, Keni; et al. (2006). "Transcription Profile Analyses Identify Genes and Pathways Central to Root Cap Functions in Maize". Plant Molecular Biology. 60 (3): 343–363. doi:10.1007/s11103-005-4209-4. PMID 16514559.  Cite uses deprecated parameter |coauthors= (help); Unknown parameter |month= ignored (help)
As OCLC/ISSN are irrelevant to the article, and the publisher is usually omitted in journal citations, as the publisher does not usually have editorial control over the content of the journal. Headbomb {talk / contribs / physics / books} 21:18, 11 August 2011 (UTC)
I agree with omitting OCLC/ISSN/publisher in this case. Perhaps in cases where there is some ambiguity about which journal is intended among several with similar names, they might be useful, but that's not true in this case because we already have unambiguous identifiers to the paper itself (the DOI and PMID). —David Eppstein (talk) 21:22, 11 August 2011 (UTC)
The problem is not with the template (though I do wonder what use OCLC is in it) but with the insistence of misinformed users with filling every single element they can, no matter how pointless it is. Circéus (talk) 22:33, 11 August 2011 (UTC)
yes and no. For many old serials which ceased publication prior to the advent of ISSNs, the OCLCN is the best ID for the serial. For the modern example cited, however, the OCLCN is redundant once the more specific ID is known. The template should handle both kinds of usage in such a way as to best help the reader looking for the source. Keep in mind too that some oclcns point to specific articles or serial volumes. LeadSongDog come howl! 03:19, 12 August 2011 (UTC)
It seems to me that I should include the ISSN (or OCLC), because by clicking on the ISSN (or OCLC) number the reader is shown a list of libraries where he can find the journal. That's not available via the doi or PMID. And I gather from the above comments that you all agree that if I include the ISSN, there is no point in including the OCLC. Wideangle (talk) 21:43, 22 August 2011 (UTC)
No, that is an incorrect generalization. I'm not aware of any case where the OCLC number is less specific than the ISSN, but it is sometimes more specific. In some cases, an OCLC number indicates a specific issue or even a specific article. For rare (physical form) publications, the ISSN will often send you to a library that has an incomplete portion of a serial (missing issues or volumes). The more-specific OCLC numbers can help avoid this. Hence, the ISSN is not a replacement for the OCLC number in the general case. LeadSongDog come howl! 15:00, 23 August 2011 (UTC)
Therefore, if I know both the ISSN and the OCLC, I should cite the OCLC. And it is not necessary that I cite both the ISSN and the OCLC. Correct? Wideangle (talk) 17:32, 26 August 2011 (UTC)
It would be better if you didn't include any of the ISSN and OCLC, as they add very very little and mostly clutter up the citation. Find a jstor, doi, OSTI id, or other article identifiers, and failing that, a url to some uploaded version (free > non-free). Headbomb {talk / contribs / physics / books} 17:58, 26 August 2011 (UTC)
Of course that only applies where you can find a doi etc - If you are using say Smash Hits as a reference, that mnay prove difficult.Nigel Ish (talk) 18:05, 26 August 2011 (UTC)
I think Headbomb is referring to articles that are available online, and for those I would agree, if it were not for the fact that people do cite paper sources. For those, the most specific identifier we can find should be cited, in keeping with wp:SAYWHEREYOUGOTIT. Other things being equal, we should prefer an identifier for the page > the article > the issue > the volume > the serial > no identifier at all. An ID that links to a free-for-everyone is nice to have, but that doesn't always exist. Broadly available is better than scarce. Scarce is better than non-existent. The scarcer the source is, the more we need details to help lay eyes on it. See the way things work at Wikipedia:WikiProject Resource Exchange/Resource Request to see how this plays out from day to day. LeadSongDog come howl! 18:30, 26 August 2011 (UTC)
OCLC numbers are potentially more specific, but they're not a formal standard identifier in the way that an ISBN or ISSN is, and so they're a lot less reliable. As they're retroactively assigned, they're not printed or recorded on the physical item (as all ISBNs and most LCCNs are), which means that any database record not directly taken from OCLC is unlikely to include them (you have to specifically search them out); any library not using OCLC services will thus return blank results for a search using OCLC control numbers. It's nice to have, if you feel a burning desire to include it, but it's definitely a bad idea to drop a standard identifier in favour of a limited-use one. Shimgray | talk | 22:55, 26 August 2011 (UTC)
"A lot less reliable"? I'm not sure I follow. The retroactive assignment is what makes them reliable: they can be created for any record, by any member library, no matter how old that record is (unlike the ISSN). While there can be redundant assignment of numbers, the duplicates eventually resolve in a predictable fashion without information loss. Can you explain the issue, perhaps with examples? LeadSongDog come howl! 03:05, 27 August 2011 (UTC)
By "reliable", I'm thinking of the odds that given an identifier, you can resolve it into a copy of the resource that is useful to you. The key phrase is "any member library"; my understanding of the way WorldCat works is that unless you create / download records with their system, your local catalogue won't have an OCLC ID in it. If you're a member who does local cataloguing and uploads, there won't be a local copy of the OCLC ID - it's treated as a Worldcat-internal thing - and if you're not a member, you're unlikely to ever care about them. (Hypothetically, you could add OCLC numbers to your own records, but a) this is a lot of effort for very little benefit, since it's very rarely searched on, and b) OCLC have in the past been involved in legal silliness which served to dissuade this.) This is the eternal problem with retroactively-assigned identifiers - unless you're aggressively trying to seek them out, they simply won't get attached to all the records reflecting that item.
So, if you search with an OCLC ID in WorldCat, you'll get a lot of hits. But imagine the copy that's actually useful to you - the one physically nearest - belongs to a non-OCLC library. No result. Or you suspect a particular library might have it, so you go to their catalogue and search locally. No result. Unless the reader actually knows in advance the restrictions of the OCLC ID, they're likely to believe either or both of these mean "it doesn't exist"... Shimgray | talk | 19:53, 27 August 2011 (UTC)
  • Also, on the initial point that "clicking on the resultant links yields the identical same WorldCat page", one question might be why ISSN sends you to WorldCat anyway! It seems a bit odd that it goes directly to one site but ISBN gives you a plethora of choices... Shimgray | talk | 22:59, 26 August 2011 (UTC)
(Reply to Headbomb/LeadSongDog) - My point about not every reference having a doi etc is that we need to careful about stripping things like ISSNs from references willy-nilly without actually looking at the reference, particularly bots. If we end up recommendinhg remoiving things like ISSNs, we need to be very clear about when it is sensible.Nigel Ish (talk) 09:28, 27 August 2011 (UTC)
And that recommendation should be ISSNs should "almost always never" be present. They are very very rarely required to identify the journal/periodical. It might be useful in very limited cases two journals having the same name, or some out of print foreign-language journal with inconsistent transliterations, but those are extremely rare situation. Not saying bots should remove them, but they certainly shouldn't be adding them. Headbomb {talk / contribs / physics / books} 23:30, 27 August 2011 (UTC)
It seems to me that I should include the ISSN (or OCLC), because by clicking on the ISSN (or OCLC) number the reader is shown a list of libraries where he can find the journal. How else can the reader get such a list? Wideangle (talk) 21:22, 30 August 2011 (UTC)
Those list of libraries are most often completely useless. If I want to access a physical copy, I'll check my library's website, or I'll go on location. If they don't have it, they'll know the best (cheapest, fastest) place to get it from. That McGill University holds a rare issue of some magazine is completely useless to everyone living outside of Montreal, and those in Montreal can check whichever library is most convenient for them (which they would probably have to do anyway, since the world cat listings are unimaginably incomplete). Headbomb {talk / contribs / physics / books} 21:38, 30 August 2011 (UTC)
"Completely useless"? A quick read of McGill University Library just lead me to this page describing their digitize-on-demand service, available to the general public for $10 per public domain volume scanned. No need to be anywhere near Montreal. Other libraries have similar services. LeadSongDog come howl! 05:02, 31 August 2011 (UTC)
It's also of extremely limited value, considered the work needs to be free of copyrights for this option to be available. Headbomb {talk / contribs / physics / books} 07:13, 31 August 2011 (UTC)

RFC on identifiers

There is an RFC on the addition of identifier links to citations by bots. Please comment. Headbomb {talk / contribs / physics / books} 15:48, 15 August 2011 (UTC)

Unfortunate interaction between template and wikiproject

Discussion consolidated at Wikipedia talk:WikiProject Academic Journals/Journals cited by Wikipedia#Unfortunate interaction between template and wikiproject
Extended content

{{Cite magazine}} and {{Cite magazine article}} redirect to Template:Cite journal.

The problem is that using {{Cite journal}} puts the article on Wikipedia:WikiProject Academic Journals/Journals cited by Wikipedia.

This puts publications that are not academic journals on the list of academic journals. Examples include: ABC Soaps In Depth, Amazing Heroes, Golden Boy Promotions, Golf World, The Amazing Pudding, Time for Kids, Toons at War, and Xtra!

In addition, this loops back and puts a entry to the list of academic journals in the "what links here list" for every article that uses {{Cite journal}}.

Possible solutions:

We could stop calling every magazine a journal and have a {{Cite magazine}} template.

We could rename Academic Journals cited by Wikipedia to reflect the fact that many of the entries are not academic journals.

(This was posted to Template talk:Cite journal, Wikipedia talk:WikiProject Academic Journals/Journals cited by Wikipedia and Wikipedia talk:WikiProject Academic Journals. I suggest centralizing the discussion at Wikipedia talk:WikiProject Academic Journals)See below. --Guy Macon (talk) 14:45, 25 September 2011 (UTC)

How is that page populated? ---— Gadget850 (Ed) talk 15:22, 25 September 2011 (UTC)
This discussion is fragmented: you, me and Headbomb have all posted in different places. HB has decided that Wikipedia talk:WikiProject Academic Journals/Journals cited by Wikipedia#Unfortunate interaction between template and wikiproject is to be the place. --Redrose64 (talk) 15:33, 25 September 2011 (UTC)
Replying there. ---— Gadget850 (Ed) talk 15:41, 25 September 2011 (UTC)
I don't agree that he should have moved the discussion to a place other than where I suggested above, but it's just a minor annoyance, not worth undoing. --Guy Macon (talk) 15:55, 25 September 2011 (UTC)

Trailing period after quote

I have just been using this template (actually {{Cite magazine}}) and noticed that the template adds a period after a quote. Unfortunately, most quotes end with a punctuation mark, so you end up with ... some quoted text.". and this looks naff. I notice that {{Cite book}} and {{Cite web}} do not add a trailing period after a quote, and look much the better for it. Could we please lose this trailing period. HairyWombat 04:43, 2 October 2011 (UTC)

That is done by:
|PS = {{#if:{{{quote|}}}||{{{postscript|.}}}}}
We need to consider that there are a myriad of uses where the editor did not include a period because the template added it. We should also consider other templates in this series. ---— Gadget850 (Ed) talk 11:16, 2 October 2011 (UTC)
Of the 16 templates in Citation Style 1, only this one and {{Cite encyclopedia}} have a default period for the quote. ---— Gadget850 (Ed) talk 12:36, 2 October 2011 (UTC)
Well, other templates in the series clearly push for no trailing period. With regard to the "myriad of uses" I see two possibilities. One is to search all 197,000 transclusions of {{Cite journal/magazine/etc}} (nine redirects in all) to find where there are non-blank quotes not ending in either ." or ?" . The other possibility is to just do it. I favour the latter. HairyWombat 16:56, 2 October 2011 (UTC)
If we don't get any objections by the end of the week, I will make the update. Ditto for encyclopedia. ---— Gadget850 (Ed) talk 12:54, 3 October 2011 (UTC)

Yes check.svg Done ---— Gadget850 (Ed) talk 18:06, 11 October 2011 (UTC)

Editor field

The editor field seems misplaced within {{Cite journal}}. Consider for instance:

{{cite journal 
|authors=Platt, John; [[Nello Cristianini|Cristianini, N.]]; and [[John Shawe-Taylor|Shawe-Taylor, J.]] 
|title=Large margin DAGs for multiclass classification 
|journal=Advances in Neural Information Processing Systems 
|publisher=MIT Press 
|editor=Solla, Sara A.; Leen, Todd K.; and Müller, Klaus-Robert 
|pages=547–553 }}

Which yields: Platt, John; Cristianini, N.; and Shawe-Taylor, J. (2000). Solla, Sara A.; Leen, Todd K.; and Müller, Klaus-Robert, ed. "Large margin DAGs for multiclass classification" (PDF). Advances in Neural Information Processing Systems. MIT Press: 547–553. 

One expects the editor field between the title and journal, not right before the title. Can this be fixed by someone? Urhixidur (talk) 22:11, 13 February 2012 (UTC)

Use |chapter= instead of |title=:

{{cite journal |authors=Platt, John; [[Nello Cristianini|Cristianini, N.]]; and [[John Shawe-Taylor|Shawe-Taylor, J.]] |chapter=Large margin DAGs for multiclass classification |journal=Advances in Neural Information Processing Systems |publisher=MIT Press |year=2000 |url=http://www.wisdom.weizmann.ac.il/~bagon/CVspring07/files/DAGSVM.pdf |editor=Solla, Sara A.; Leen, Todd K.; and Müller, Klaus-Robert |pages=547–553 }}

Platt, John; Cristianini, N.; and Shawe-Taylor, J. (2000). Solla, Sara A.; Leen, Todd K.; and Müller, Klaus-Robert, ed. Advances in Neural Information Processing Systems (PDF). MIT Press: 547–553 http://www.wisdom.weizmann.ac.il/~bagon/CVspring07/files/DAGSVM.pdf.  Missing or empty |title= (help); |chapter= ignored (help)

Or; using the first and last parameters:

{{cite journal |last1=Platt |first1=John |last2=Cristianini |first2=Nello |authorlink2=Nello Cristianini |last3=Shawe-Taylor |first3=John  |authorlink3= John Shawe-Taylor |chapter=Large margin DAGs for multiclass classification |journal=Advances in Neural Information Processing Systems |publisher=MIT Press |year=2000 |url=http://www.wisdom.weizmann.ac.il/~bagon/CVspring07/files/DAGSVM.pdf |editor1-last=Solla |editor1-first=Sara A. |editor2-last=Leen |editor2-first=Todd K |editor3-last=Müller |editor3-first=Klaus-Robert |pages=547–553}}

Platt, John; Cristianini, Nello; Shawe-Taylor, John (2000). Solla, Sara A.; Leen, Todd K; Müller, Klaus-Robert, eds. Advances in Neural Information Processing Systems (PDF). MIT Press: 547–553 http://www.wisdom.weizmann.ac.il/~bagon/CVspring07/files/DAGSVM.pdf.  Missing or empty |title= (help); |chapter= ignored (help)

I may have introduced that extra period in a recent fix— I will look into it. ---— Gadget850 (Ed) talk 23:02, 13 February 2012 (UTC)

Not my fault after all. I proposed a fix in {{citation/core}}. ---— Gadget850 (Ed) talk 15:50, 14 February 2012 (UTC)
Upon reflection, |chapter= is formatted in italics, which is not right. ---— Gadget850 (Ed) talk 13:45, 14 February 2012 (UTC)

Thanks for the tip. Although I think the real answer was "Use 'cite book' instead, swapping 'journal' into 'title'". :-) Here's a comparison of calls:

{{cite journal 
|authors=Platt, John; [[Nello Cristianini|Cristianini, N.]]; and [[John Shawe-Taylor|Shawe-Taylor, J.]] 
|chapter=Large margin DAGs for multiclass classification 
|journal=Advances in Neural Information Processing Systems 
|publisher=MIT Press 
|editors=Solla, Sara A.; Leen, Todd K.; and Müller, Klaus-Robert; eds 
|pages=pp. 547–553 }}
{{cite book 
|authors=Platt, John; [[Nello Cristianini|Cristianini, N.]]; and [[John Shawe-Taylor|Shawe-Taylor, J.]] 
|chapter=Large margin DAGs for multiclass classification 
|title=Advances in Neural Information Processing Systems 
|publisher=MIT Press 
|editors=Solla, Sara A.; Leen, Todd K.; and Müller, Klaus-Robert; eds
|pages=547–553 }}

Platt, John; Cristianini, N.; and Shawe-Taylor, J. (2000). Solla, Sara A.; Leen, Todd K.; and Müller, Klaus-Robert; eds, eds. Advances in Neural Information Processing Systems (PDF). MIT Press: pp. 547–553 http://www.wisdom.weizmann.ac.il/~bagon/CVspring07/files/DAGSVM.pdf.  Missing or empty |title= (help); |chapter= ignored (help)

Platt, John; Cristianini, N.; and Shawe-Taylor, J. (2000). "Large margin DAGs for multiclass classification". In Solla, Sara A.; Leen, Todd K.; and Müller, Klaus-Robert; eds. Advances in Neural Information Processing Systems (PDF). MIT Press. pp. 547–553. 

That trailing period is annoying, I agree. But note also the differing behaviour of the pages fields. There seems to be no way to get the quoted title field of 'cite journal' to appear in a 'cite book' (it would be before the latter's italicised title). Urhixidur (talk) 20:34, 17 February 2012 (UTC)

The output of |chapter= absolutely should not be italicized. I have a whole bookshelf of style guides, and none of them recommend doing that. But that's a citation/core problem. For now, the way around this problem if you stick with {{cite journal}} is probably creative use of |at= which doesn't impose any formatting. The long term fix is clearly, as proposed, to ensure that the editor field comes between the paper title and the journal name, because it applies to the journal, not the paper. Anyone needing to be cited as an editor of a paper can be done like so: |last=Garcia |first=Dolores M. (ed.). For this particular case, you probably wanted to use {{cite book}} all along, not {{cite journal}}, since it's, um, citing a book, not a journal. :-) — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 09:38, 5 March 2012 (UTC)


 Request withdrawn

The markup to handle page and pages is currently:

  |At = {{
          #if: {{{journal|{{{periodical|{{{magazine|{{{work|}}}}}}}}}}}}
             #if: {{{page|}}}
             |p. {{{page}}}
                #if: {{{pages|}}}
                |pp. {{{pages}}}

Thus, if journal or one of the aliases is defined, then the p. and pp. abbreviations are not added. Does anyone know why this is desired? ---— Gadget850 (Ed) talk 13:58, 27 February 2012 (UTC)

This is one of those annoying inconsistencies between the citation templates than needs to be eliminated. And there's certainly no reason to tie whether or not to display "p."/"pp." before page number(s), to whether or not the work has properly been specified. That's like deciding whether or not to pay your electric bill based on how cloudy it was yesterday or how good your coffee turned out this morning. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 09:17, 5 March 2012 (UTC)
Since this markup is in each CS1 template, I have made a centralized proposal at Help talk:Citation Style 1#Page/pages with journal. ---— Gadget850 (Ed) talk 15:26, 5 March 2012 (UTC)

Article and abstract need separate URL parameters

We need to differentiate between links to full article copies (free or not) and abstracts. This is very important for verifiability and original research checking purposes. It is a very common problem that editors, especially those who are not experts in the field in question, come to incorrect (or even arguably correct but novel-synthesis) assumptions about the meaning of a full paper based on a highly compressed abstracts, then add their synthesized version of the facts to an article, with citations that look complete and valid. Help talk:Citation style 1 has already seen a lot of debate about how misleading it can be to link to abstracts, snippet views and other non-full texts at all. At least no confusing them with links to full texts would be a start toward resolving those disputes.

I propose that we have the display change to something like the following pseudocode mockup "[{{{url|}}} {{{title}}}}]" ([{{{abstract-url|}}} abstract]) (there would really need to be some "ifs" in there to not display the "[...]" and "([...])" punctuation if not needed, of course). This could actually be added to the Template:Citation/core meta-template as something available for any template. I can see it being very useful aliased as |review-url= for {{cite book}}, {{cite video}}, {{cite comic}}, etc. I'm raising it here first because it's actually crucial here, versus just a nice extra feature for the other citation templates. Actually implementing it would need to be a proposal at Template talk:Citation/core. I'd like to go there with a consensus already arrived at here that we need this.

I'm aware that this would not do anything to fix the innumerable cases of |url= already linking to abstracts instead of articles. Only human review and cleanup over time, as part of source verification (e.g. in WP:GAN, WP:FAC and other peer-reviewing processes as well as day-to-day article cleanup) could possibly address that. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 09:13, 5 March 2012 (UTC)

I have had the same thought, and there have been some discussion that have touched on this. We have doi, PMC, JSTOR and other identifiers that link to abstracts, so a generic abstract field that allows a URL would fit right in. laysummary is for summaries and reviews, but I am not enthused about the way it is presented. ---— Gadget850 (Ed) talk 11:43, 5 March 2012 (UTC)
Right, I'd forgotten about that. I think the concept of an abstract or lay summary could be merged, and something like what is presently laysummary should be reserved for reviews (the two would often apply to a single cited work). — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 13:04, 5 March 2012 (UTC)
This one needs to go to {{citation/core}}. ---— Gadget850 (Ed) talk 14:54, 5 March 2012 (UTC)
I for one really don't see a need for a distinct "abstract" link. In fact, I'm rather against it. Very concerned with the additional clutter it'll add to citations.Headbomb {talk / contribs / physics / books} 17:49, 5 March 2012 (UTC)
We already have a precedent for linking to abstracts via doi, Bibcode, JSTOR and the like. Where do you suggest the link goes? Do we just delete it from {{URL}}? What if an abstract link is sneaked in through |id=? ---— Gadget850 (Ed) talk 17:58, 5 March 2012 (UTC)
doi, bibcode, etc... all link to the article, or certain database entries about the article (who cites it, similar articles, abstracts, free versions, etc...), etc... That they very often happen to have abstracts is besides the point, and also make some explicit "abstract" link both redundant and useless. If you have a free link to the article, you don't need the abstract link. If you don't have a free link to the article, then use |url= (unless of course, it's redundant with a doi or whatever). Headbomb {talk / contribs / physics / books} 18:04, 5 March 2012 (UTC)

I think (1) there is a big difference between links that go to an abstract page from which the full paper can be found, vs links that only go to an abstract without any way to access the full paper, (2) there is not much point in supplying url= links for abstract-only pages that are not part of some big unified database like pubmed or whatever (which have their own separate parameters already), (3) there is also not much point in distinguishing abstract+link to full paper lurls from links that go directly to the full paper. So I don't see the use case for the proposed new parameter. As Headbomb says, it just seems like clutter. —David Eppstein (talk) 18:05, 5 March 2012 (UTC)

Then how would you resolve the issue that people very frequently misleadingly link to abstracts as if they were full papers? Not every linkable abstract is in database like PubMed for which we have special code? — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 05:31, 6 March 2012 (UTC)
You have examples of this practice? Because I don't recall seeing it myself. In any case linking to an abstract is only weak evidence that the person doing the linking hasn't really read the paper; there are other ways of reading papers than via the internet. —David Eppstein (talk) 05:48, 6 March 2012 (UTC)
Do you think I would lie about it? How about virtually every single case of a citation to a journal article involving a journal that is not available for free online and in which there is a |url= parameter? (About the only exceptions are when the writer of the article posts it on their own personal website or when someone posts a copy of it elsewhere without permission, which is a copyvio we can't link to.) My usual technique for dealing with it is to remove the URL parameter, and after {{cite journal|...}} but before </ref> add [pasted the URL here Abstract available from name of publisher of abstract]., but this gets very tedious, which is why I suggest a parameter to automate it. I never said anything about evidence about whether the citer read the entire article; this is about not misleading fact-checkers into thinking that a citation is complete and easily verifiable, which may not be at all. It would clearly indicate which sources need more in-depth checking, e.g. by someone who has an account at a university and has free access to journals through the school's e-library system, a privilege unavailable to most of us. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 08:16, 6 March 2012 (UTC)
I'm against this. It isn't needed anyway as there are few reliable sources that aren't indexed by one of our existing parameter-databases that have abstracts such as PMID. People link the abstract in the url for several reasons. Sometimes the abstract is all they read. Sometimes they read the paper, link to its url but nearly everyone else just gets the abstract or worse, a page demanding $30 for two sheets of A4. Having an extra parameter will only encourage folk to fill it in, which nearly always isn't desirable. We have too many hyperlinks in our citations as it is and every extra hyperlink is one more to rot. Colin°Talk 09:01, 6 March 2012 (UTC)

Need for journal abbrevation parameter

The template needs a journal abbreviation parameter. Why? Ironically because journal abbreviations are potentially confusing, especially as used by secondary and tertiary sources. This parameter would be especially useful for managing references of scientific journals from previous centuries. When editing an article, sometimes you wish to cite an original source that is mentioned by a non-primary source. The source may only use a journal abbreviations for the citation. There are several problems here because finding the actual journal to which the abbreviation refers may not prove possible, or simply be too time-consuming. For example, you may not find any journal that matched or you may find more than one journal that is a potential match but be unable to resolve the ambiguity. Ambiguity arises when the journal is inaccessible online and you cannot see the original article to confirm the match. This is true of many old, discontinued publications. It is also true of some modern publications behind paywalls. In the end, you may be left with doubt about the full name of the journal for a reference you want to include. (This assumes you trust the source that produced the ambiguous citation in the first place.)

How should it work? I see these possible scenarios:

  1. The "journal" parameter trumps a new "journal_abbr" parameter. If just the "journal" parameter is used by a citation, its text will be used in the reference listing. If just the "journal_abbr" parameter is used, its text will be used. If both are used, the "journal" parameter takes precedence. The disadvantage is that full journal names can be long and they might clutter-up the references text. The advantage of this is that the ambiguity of journal abbreviations used in the original sources is avoided.
  2. The "journal_abbr" parameter trumps the "journal" parameter. This is mostly the opposite of the previous case. A new disadvantage arises from the potential of multiple abbreviations being used to refer to the same journal. I really like to maintain a reference as exact to its nature as possible, so for instance, if some source uses Bull. Soc. ent. Fr. to refer to Bulletin de la Société Entomologique de France while another uses Bull. Soc. Entom. Fr., I think the "journal_abbr" tag should be used to use the abbreviation actually used by the source.
  3. Both tags are treated as equals. If just "journal" is used, only it is displayed in the reference. If only "journal_abbr" is used, only it is displayed in the reference. If both are used, both are display with the journal abbreviation in parentheesis. For example, Bulletin de la Société Entomologique de France (Bull. Soc. Entom. Fr.).

A side benefit arises. Some journals have had the same name but different abbreviations. This would help resolve this citation problem.

I suppose that I support the "treated as equals" solution. Jason Quinn (talk) 06:10, 13 March 2012 (UTC)

Seems like more useless clutter to me. How some reference is cited by some work is should not affect how we cite that reference. Headbomb {talk / contribs / physics / books} 08:14, 13 March 2012 (UTC)
I see this as equivalent to saying that citations that come from a non-primary should always use exactly the same form as the non-primary source used. Why? Because as I explained, there are situations where it's not 100% clear to what the citation is referring. Doing nothing is fine if this is considered to be too small a problem to worry about. In practice, however, I have found this desirable now and again. Jason Quinn (talk) 08:23, 13 March 2012 (UTC)
I know various disciplines use abbreviated journal names, but Wikipedia is an encyclopedia. Given our target audience, there is no reason to use such abbreviations. ---— Gadget850 (Ed) talk 09:12, 13 March 2012 (UTC)
We ask that the various references within an article be mutually consistent as regards style, but we don't mandate what that style should be. If we suggest that refs to journals should use the style of that journal, then unless every ref is drawn from one journal, we will soon see inconsistency of style. This is not to say that the styles used by journals are wrong: if an article has several refs to e.g. Journal of Physics, and those refs are formatted in the style preferred by that journal, this is OK provided that all other refs in the same article are also formatted in the style of Journal of Physics. --Redrose64 (talk) 14:22, 13 March 2012 (UTC)
Journal abbreviations can and do get used. They are unavoidable in cases where it's the only information linking to an unidentified original source. That is one of my main concerns but its subtle. If you don't think it's an issue, try adding a bunch of references to a handful of obscure insect articles when you're not an expert on insects. You will find yourself spending a large amount of time trying to figure out what journals are actually being cited by non-primary sources because of abbreviations. Even when you make progress, you will often find yourself only partially sure you have found the correct match. Naturealist journals have this tendency to change names and have very similar names. You can easily find yourself spending 10 or minutes or more just trying to resolve the full journal's name. It's a logistic roadblock. It also ends up putting more responsibility on readers and editors and expects them to know if "Ent. Journ." and "Entom. J." refers to the same journal or not from the 1870's. When somebody's gone to the length of actually resolving the ambiguity and invested the time to do it, it'd be good to preserve that effort while at the same time preserving the original citation style (otherwise it makes the reader wonder if the citation in the non-primary source was referenced incorrectly). Also I think RedRose64's point about referencing style is a good one. Style is largely up to the editors as long as it's consistent. Adding a new "journal_abbr" parameter would allow people to build upon this template to include new styles that are in common use. Considering many fields such as physics primarily use abbreviations, this is a nice capability. Jason Quinn (talk) 15:58, 13 March 2012 (UTC)
There shouldn't be any "unidentified original source". This isn't how Wikipedia refs work. If you find a fact in Source A, the Wikipedia article should reference Source A, even if Source A references Journal B. You can only reference Journal B if you happen to have a copy of Journal B, and have verified that it supports the fact under consideration: in such a case there is no need to use abbreviations - write out the name of Journal B in full as it appears on the cover (or contents page) of the copy that's in front of you. --Redrose64 (talk) 17:08, 13 March 2012 (UTC)
I don't entirely agree. If source A says that some fact in the article was first published in source B, then I think it's ok to repeat the same statement, as long as the reference to source B is clearly marked as coming from source A (the wording I use is to put "As cited by source A" immediately following the reference to source B). However, this still doesn't justify the need for two different journal name parameters. If you know the full name for the journal in source B, you can use it, and if you only know some abbreviation then just use that — why does it need to be in a differently named parameter? —David Eppstein (talk) 17:21, 13 March 2012 (UTC)
The purpose of a citation is to provide a source to validate content while providing enough information to identify the source so others can find it. As an encyclopedia, we need to make these sources accessible to a very general audience; this includes avoiding jargon such as abbreviations. This discussion has broader implications and should be taken to Wikipedia:Citing sources.
The example of "Ent. Journ." and "Entom. J." is puzzling. If you actually read the source, then you should know the full name of the journal. ---— Gadget850 (Ed) talk 17:40, 13 March 2012 (UTC)
I agree, it should be possible to validate content... which often means using a primary source for reference. If you reference a secondary source and that source's citation to the supposedly primary source is a online search dead-end, have you validated content by citing the secondary source? Not really. I've noticed an observation fact that secondary sources often use primary sources in questionable ways. I've also noticed that the secondary sources often flub the citation to the primary source, getting things like the year wrong. This is why I often try to trace statement back to original sources. Regarding avoiding abbreviations because they are "jargon", I think you misunderstand my point of view. I am not encouraging the use of abbreviations over the full names, in fact, exactly the opposite. Full names are better. I want people to use the full name of journals in the reference. Often editors just don't bother to give the full name, another fact. Abbreviations are in widespread use at Wikipedia like it or not. When used properly my proposal helps readers track down sources and dispel ambiguity. It is not meant to encourage the lazy use of abbreviations just to save typing. Regardless, you are making a judgment call about citation style that isn't a policy or a guideline (to my knowledge). If editors would prefer to use exclusively abbreviations in citations for an article, that's their prerogative. All this said, Redrose64's argument that I should rely on the secondary source instead of the primary when I cannot confirm the primary is a strong one. It may force me to reconsider my argument. It does not invalidate my proposal, some benefits I've suggested have gone unaddressed, but just undermines the main argument I've used so far. There are practical issues though that go against Redrose's point of view. For instance, species articles very commonly use references to sources where the reference itself came from an online database. (This situation was a motivating one for my entire proposal.) According to Redrose64, all those citations should actually be citing the database itself. Another solution would be for the citations themselves should be cited, which would be silly. Either of these two solutions would be clunky and impractical and I would hope that would be acknowledged even if my proposal were declined. Jason Quinn (talk) 18:14, 13 March 2012 (UTC)

Proposed Parameter: cnki.

I noticed relatively recently that there are cite journal entries with the doi parameter starting with CNKI such as at Ch (computer programming). the doi parameter should have the IDF parameters that start with 10. , but these don't. These DOI's are Chinese, run through cnki.com.cn (what we would want to link to is at en.cnki.com.cn if it existed, www.cnki.com.cn if it doesn't, I think). As I see it there are three steps with CNKIs, the first is to simply have it as a printed parameter in cite journal, the second would be to figure out how to generate the URL at cnki.com.cn (which appears to be relatively simple to generate) ( CNKI:SUN:WOLF.0.2009-05-018 gives http://en.cnki.com.cn/Article_en/CJFDTOTAL-WOLF200905018.htm , but I'm still researching) and the third would be to create a cite cnki template which is automatically filled by a bot in the same way as cite doi and cite pmc. I'd love opinions on this, I haven't tried to add a parameter to a major template before.Naraht (talk) 13:23, 14 March 2012 (UTC)

For background to this, see Template talk:Cite doi#*Chinese*(?) DOI?. --Redrose64 (talk) 14:00, 14 March 2012 (UTC)
If these are only going to occur rarely, they should just be put under the |id= parameter rather than adding a separate parameter for them. —David Eppstein (talk) 14:57, 14 March 2012 (UTC)
For the number of articles that Wikipedia has, what is rarely, ten, a hundred, a thousand?Naraht (talk) 14:59, 14 March 2012 (UTC)
The last time we did some identifier overhaul, I think the cut-off was more or less ~100 articles. It might also be time to add |id={{hdl|foobar}} as |hdl=foobar. Headbomb {talk / contribs / physics / books} 15:55, 14 March 2012 (UTC)
I think in this case, both the ones that were in cite journal as a doi starting with CNKI as well as those current links to cnki.com.cn may be appropriate (Also, in terms of links is it possible to do something like if en.cnki.com.cn/blahblah doesn't exist, do www.cnki.com.cn (The English as a link being preferred to Chinese))Naraht (talk) 16:01, 14 March 2012 (UTC)
Please see the outline at Template talk:Cite doi#*Chinese*(?) DOI?. This needs to be added to /identifier and /core before propagating to the cite templates. ---— Gadget850 (Ed) talk 17:05, 14 March 2012 (UTC)

The volume, issue and page number output is non-encyclopedically geeky

Output in the form "4 (9): 6" needs to be changed to "Vol. 4, No. 9, p. 6" per WP:COMMONSENSE (and Wikipedia is not a journal), and to match the other citation templates, which use plain English instead of gobbledegook. There's no reason for this template to use an excessively clipped and geeky format that no one except professional academics and their grad students understands. It's not even universally used in academic publishing, but only recommended by ABA and CBE/CSE style guides. Both Chicago (our #1 go-to guide, at WP:MOS) and Turabian call for "no.". Several that don't [and note that they are all for entirely academic contexts], including APA, MLA, Oxford/Hart's, and MHRA, do not use "4 (9)" formatting, but some other variant, like "4.9", "4 9", "4, 9" etc. Insisting on one particular minor variant of the most reader-unhelpful possible way to convey volume and issue information in WP is downright "user-hateful". It's just another case of WP:ILIKEIT and WP:SPECIALSTYLE in favor of professors and scientists against the interests of the vast majority of the readership. (PS: I say that as someone with a science degree.) — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 13:00, 5 March 2012 (UTC)

I agree. There has also been a lot of discussion over bolding the volume. The volume and issue formatting is in {{citation/core}}, and the page number issue is in every template. ---— Gadget850 (Ed) talk 13:06, 5 March 2012 (UTC)
I know, but Template talk:Citation/core is unlikely to make a change that only affects {{cite journal}} output without seeing a discussion about it here. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 05:47, 6 March 2012 (UTC)
Please don't use citation template talkpages as a surrogate for MOS talkpages. If there is consensus at MOS for preferring a specific style, all that's needed here is a link to that discussion. LeadSongDog come howl! 15:00, 5 March 2012 (UTC)
There is no MOS page for citation templates. There is no MOS guideline for the presentation style of locators and identifiers in a citation, nor for the presentation style of a journal citation. The template talk pages have a long history of discussion and implementation of the presentation for these templates. If I am incorrect, please point me to the proper point of discussion. I am attempting to centralize these discussions at Help talk:Citation Style 1, but that is still ongoing. ---— Gadget850 (Ed) talk 15:32, 5 March 2012 (UTC)
Sorry, Gadget. I seem to have misremembered, unless a page has been deleted. The closest I can find are Wikipedia:Citing_sources/Example_edits_for_different_methods#Footnotes and Wikipedia:CITEVAR. The thing is we still have many articles with freehand wikicoded citations. If the templates implement a change of something that diverges from the established rendered style, pages which have both templated and freehanded citations will get mixed results. As a result something as trivial as removing the bolding on the issue presentation by the templates would uglify articles all across the project. We need to have a unified discussion page for a style of citation, not just for templates that implement it. LeadSongDog come howl! 15:23, 6 March 2012 (UTC)
No kidding. Let me rephrase: I've yet to see anyone anywhere make a credible case for using the confusing, geeky and very poorly reliable-source-supported format "4 (9): 6" (only two "authorities" of any note recommend it, and at least 6 major competitors do not; of those; two uniformly prefer what I propose, and all the rest just make up their own variants like "4.9", "4 9", etc.). Using plain English instead of academic-nerd gibberish makes basic common sense and actually helps readers. Is that clearer? I'm unaware of any policy that would require the particular pile of regulars at WT:MOS to be involved. This talk page emphatically is the correct venue for discussion of this template's output, regardless where the code is, and regardless of whether MOS could conceivably eventually decide to codify a citation template output style guideline. I'd actually be strongly in favor of one. I'm also strongly in favor of a manned mission to Mars, but this does not stop me going about life in normal, practical ways pending that miracle. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 05:47, 6 March 2012 (UTC)

One main argument not to use "plain English" is that the terminology varies from journal to journal, from Volume, Issue, Number, etc. It gets worse with "New Series", "Old Series", etc. So forcing a plain English naming of "issue" for the secondary document number identifier won't work when Fred's Big Plane Journal uses "number." etc. This would require us to provide fields for non-automatic number-type identifiers, or encourage plain text fields. Plain text fields will lead to variance between: "Issue 1" "Volume 1 Issue 1" "Volume 1, Issue 1" "Volume 1, issue 1" "Volume 1 number 1" etc. I'm not against better results here, but, this is an issue with a suggested change towards plain English in identifying subsections of a continuing journal title. Fifelfoo (talk) 06:15, 6 March 2012 (UTC)

"Issue" and "number" are synonymous in this context, though. "New Series" would have to be handled in |volume= anyway (e.g. |volume=IX (Old Series)), regardless of formatting, since there is no parameter for it, and it's an odd enough case we don't need one. Even aside from any such alleged number/issue problem, that doesn't mean there's any reason not to have "Vol." output by |volume= and at least a default like "No." output by |issue=, which could be overridden with something if we really, really, thought people couldn't figure out that "number" and "issue" mean the same thing. Given that "4 (9): 6" uses neither "number" nor "issue" nor anything else, I think this is clearly a non-problem. It's not like we have a principle in play that says 'it must be "4 (Issue 9): 6", or "4 (Volume 9): 6" following the indicia of the publication" that my proposal would violate. We're already ingnoring the "issue"/"number" distinction with impunity. If I were Spike Lee, I would not expect {{cite video}} to support |joint= and output something special for my film with that parameter that was somehow different from the output of |work= in that template. WP doesn't care how producers of works care to label them. Spike Lee knows, and we know, that what he calls "joints" everyone else calls "films". The producers oand readers of journals, along with our readers, know that what some periodicals call an "issue" is otherwise referred to as a "number" or "no.". Most of us probably figured that out within a few years of learning to read. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 09:20, 6 March 2012 (UTC)
Another argument is nobody cares. This is the Internet age folks. If there are urls and database keys in the citation then that's pretty much all anybody needs. The rest should be given for completeness but worrying about it at a MOS level? Don't you have a life? I next expect someone to argue the volume number should be spelled out if less than 10. Colin°Talk 09:01, 6 March 2012 (UTC) On second thoughts, make it as verbose as you can. Then, when editors are horrified that their references section is twice as long as their articles, they might abandon these blasted templates and write some compact citations by hand, free from the whims of stylists and bot tweaks. Colin°Talk 09:10, 6 March 2012 (UTC)
If nobody cares, why attack and insult me about it? Go do something productive instead. In point of fact, WP is intended to be reused in any way that anyone cares to reuse it, including print. The "there's a URL, so it will never matter" idea is an invalid argument. And the "4 (9): 6" crap is major [[WP:ACCESSIBILITY] problem, besides. No one would argue for spelling out volume numbers under 10, since it's not running prose. Adding a few characters will not result in "twice as long" citations; hyperbole is unbecoming and unhlepful. If you hate citation templates so much, don't use them. But don't try to disrupt discussions about how to format them in ways that don't confuse our readers. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 09:20, 6 March 2012 (UTC)
How many readers have complained about this issue? Colin°Talk 11:03, 6 March 2012 (UTC)
Currently we don't present false information, we present an integer or text string at a hierarchical level without declaring it to be anything. Forcing a text string will invariably introduce incorrect information. Fifelfoo (talk) 09:46, 6 March 2012 (UTC)
Somebody mentioned a lack of references for the all-numeric "volume (issue) pages" form. Students at the University of Oxford can easily get hold of this:
  • Fisher, David; Hanstock, Terry (1998). "chapter 5: Journal Article References (Harvard)". Citing References. Oxford: Blackwell. p. 10. ISBN 1 85377 992 X. 
"The order of the elements (including upper and lower case and punctuation) of the reference is:
AUTHOR, Date. Article title. Journal title, volume (part), pages.
GREENFIELD, J., 1990. The Sevso Treasure: the legal case. Apollo, 132 (341), 14-16.
...the numbers before the brackets refer to the volume and those inside the brackets refer to the issue or part number." --Redrose64 (talk) 13:53, 6 March 2012 (UTC)
Just one style guide among many, none of which agree on exactly how to do this "geeky citation style"; see User:Gadget850/Cite comparisons for examples. PS: Our readers around the world are not except in very small part "students as the University of Oxford". We have to write in a way that is understandable by everyone not just students in a partcular place. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 19:57, 19 March 2012 (UTC)
My point wasn't that we must use Oxford style. My point was that the "132 (341), 14-16" method isn't something that Wikipedia made up one day: it's out there in the wild. --Redrose64 (talk) 21:27, 19 March 2012 (UTC)
I'm also puzzled by '"New Series" would have to be handled in |volume= anyway (e.g. |volume=IX (Old Series)), regardless of formatting, since there is no parameter for it, and it's an odd enough case we don't need one'. We already have a |series= parameter for precisely this purpose; indeed, the doc page states
  • series or version: When the source is part of a series, such as a book series or a journal where the issue numbering has restarted.
There is therefore no need to misuse |volume=. --Redrose64 (talk) 14:12, 6 March 2012 (UTC)
Great! I had not noticed the addition of that parameter. But it has nothing to do with the proposal to standardize. All this does is invalidate the objection above that '"[o]ne main argument not to use "plain English" is that the terminology varies from journal to journal ... with "New Series", "Old Series", etc.' PS: You're misusing the {{xt}} template; see {{tq}} for what you want. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 19:57, 19 March 2012 (UTC)
The |series= parameter was added to {{cite journal}} in November 2008. --Redrose64 (talk) 21:13, 19 March 2012 (UTC)
I agree with SMcCandlish - the way the template presents vital information into a strange and arcane sequence of numbers and brackets with no explanation seems designed to obscure rather than explain. We should be writing references to be as clear as possible to as many readers and potential editors as possible, not just for an elite group who has been taught one particular way of presenting academic references in college or university. The cite journal template is not just used to present academic jounals, but a great range of both academic and non-academic reliable sources of greatly differing ages and accessibility (not everything will have a URL or database key), so any templates that are used as a de facto "wikipedia standard", which like it or not the cite xxx templates are, should be as clear to use and for anyone to read as possible, with things like Vol and number stated, which can only make things easier for the reader, and unneccessary abbeviations eliminated.Nigel Ish (talk) 21:48, 6 March 2012 (UTC)

Proposal to use "Vol.", "pp.", etc. in citations instead of ambiguous formatting like "9 (4): 7"

You are invited to join the discussion at Help talk:Citation Style 1#RfC: Use "Vol.", "pp.", etc. consistently between citation templates, instead of ambiguous formatting like "9 (4): 7". The talk page at Help talk:Citation style 1 is where the discussion about most of our citation templates is centralized. — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 19:53, 19 March 2012 (UTC) — SMcCandlish   Talk⇒ ɖ∘¿¤þ   Contrib. 19:53, 19 March 2012 (UTC)

The "url" parameter

Discussion about changing Template:Citation/core so that the URL given with |url= won't slip to chapter is happening at Template talk:Citation/core#The "url" parameter.Dmitrij D. Czarkoff (talk) 00:01, 21 March 2012 (UTC)

Open access