Talk:Handel flute sonatas

Update of reference and citation strategy

I propose that the referencing and citation strategy be updated as demonstrated by this edit. The use of {{cite book}}{{cite journal}} and {{sfn}} assist the reader by providing consistent formatting and linkages (via hyperlinks) within the article, and assist editors via a consistent parameter-driven approach for future updates to reference and citation syntax (e.g. via bot-assisted updates). The strategy demonstrated above reduces the amount of citation syntax within the article text – which is a bonus for editors, and it cuts down on repeated information in the References section (as demonstrated by the article state following the revert of the above edit) – which is a bonus for readers.

• Support as proposer (and article creator).GFHandel ♬  23:04, 17 January 2013 (UTC)

Gosh! That sure does sound impressive! What does it mean in English? ;-) Before you take offense, I think I'm on your side in this—I just want to make sure I understand what this means from an ordinary reader's point of view. I also reverted your edit only as a matter of procedure, since WP:CITEVAR mandates retention of the established format until consensus to change it has been achieved. Personally, I prefer using source lists and shortened in-text citations. However, there may be disadvantages to the use of citation templates, about whjch I am very wary. I would like to see a discussion based on the actual sources used here, as well as any sources likely to be added in future. Amongst the already-in-place items, I am particularly concerned about the New Grove citation, which I notice you did not convert. Also, the {{cite book}} template is not suitable for journal articles, because amongst other things it has no parameter for inclusive page numbers.—Jerome Kohl (talk) 01:29, 18 January 2013 (UTC)

I meant to write {{cite journal}} (as used in my reference-update edit). The inclusive page numbers are in the {{sfn}} template, and it isn't necessary to describe the journal article in more detail than is required for location as a source (e.g. there's no need to describe the page range of the article since the {{sfn}} page range lands the interested reader somewhere within the article). If anyone does feel the need to add the page range, perhaps the |volume parameter could be used (since it's free-form text anyway).

There is a specific template for {{NewGrove2001}}, but I don't think it's possible to add something like |ref=harv to it (for use in {{sfn}}). I might ask another knowledgeable source about that, because I am keen to see if it can be moved to the Sources section in such articles.

BTW, WP:CITEVAR's "mandate" doesn't mean that changes can't be made (I do it all the time – with extremely little push-back). I'm going to continue with the strategy that: if questioned, then talk page discussion is required – as is happening here (and I'm comfortable with that). Any other approach is simply too time-consuming and bureaucratic (and if I eventually get banned for my approach, then so be it). I feel I'm supported by real-world practice in my approach to referencing changes because I think I'm batting about 0.994 (dropped a bit today) with lack of questioning for such changes – with other editors overwhelmingly appreciating how I'm trying to improve articles, and with the use of the more modern approach to referencing (that is, thank goodness, gaining traction across WP).

Regarding this article, I stick by my (at least) four points that indicate that there are more advantages than disadvantages with the scheme I adopted to improve the article. I'm not aware of the disadvantages of the Cite templates (especially in this article); and if slower load time is the concern, then I think we're okay here – there being slightly less than the 250 or so Cite templates used in the Britney Spears article (which does have a marginally higher hit rate than an article about Handel's flute sonatas – unfortunately).:-)

GFHandel ♬  02:48, 18 January 2013 (UTC)

Let me reiterate that I agree with you in principle. The current reference format is just about the one I like the least of all: full references in footnotes. I find it especially problematic in online documents, and doubly so on a freely edited forum like Wikipedia. The greater the number of inline citations, the more troublesome it gets. That said, I'm not sure that your "modernization" goes quite far enough. Once the inline reference has in effect been reduced to a marker, with a link to a list of sources using the sfn template, I fail to see the point of putting one link in the text (a footnote number) that only leads to another one (in a footnote). Why not cut out the middleman and use the sfn template the way it was evidently designed to be used—with direct links from parenthetical referencing? That way, we would at least be updating the referencing style to the late 19th century! Problems with the templates can be discussed separately, but I think should be discussed before moving ahead with any change in the referencing format that might only have to be redone later.—Jerome Kohl (talk) 04:45, 18 January 2013 (UTC)

The bidirectional links are a result of having one described source and multiple citations to that source. I don't mind the links and I'm sure that (the one in a hundred) readers who need to follow them won't either. However I am keen to learn, so could you please give us an example of the way the sfn template was designed to be used – either by editing a case in this simple article or by directing us to an article demonstrating that usage? Thanks. GFHandel ♬  19:10, 18 January 2013 (UTC)

My apologies, I was confusing the sfn template with {{harv}} and {{harvnb}}. Obviously, the sfn template is designed for use with shortened footnotes, as its TLA might suggest. I guess what I should have said is that harv templates would be better (more "up to date") than what you are proposing.—Jerome Kohl (talk) 19:51, 18 January 2013 (UTC)

• Support – I have begun to use this style of citation more and more, as it is cleaner, easier to read inline, and easier to maintain. One point on the conversion edit: there is an issue= parameter for cite journal, so you could change, e.g. "volume=13, No. 4" to "volume=13 | issue=4". —Torchiest ^talk_edits 20:51, 18 January 2013 (UTC)
• Support and Comment: {{sfn}} are superior to {{harvnb}}, because harvnb does not automatically collate for identical references. Each citation must be named : <ref name=Hughes450>{{Harvnb|Hughes|2002|p=450}}</ref>, for example. This leads to problems with multiple citations that are not identical yet have the same name. The second and subsequent citations are masked by the first one. Citations that click straight down to the bibliography are not a good idea, as then the reader is not told what page the information is on. For journal articles, you can list the full page range in the bibliography and use the sfn template to specify the exact page number within the article. This is useful for larger journal articles. I use the sfn system all the time, and find it easy to work with and maintain. -- Dianna (talk) 21:06, 18 January 2013 (UTC)
• Comment With respect to "Citations that click straight down to the bibliography are not a good idea, as then the reader is not told what page the information is on", how does a parenthetical (i.e., Harvard) reference not tell the reader what page the information is on? The syntax is ([Author] [Year], [Page]), for both books and articles. Naturally, the page range of articles is given in the list of sources in any case. In the interests of full disclosure, I should say that I have almost no experience at all with sfn templates, and templates in general give me the willies, since I so frequently find things belonging in references that are not accommodated by them (series titles in book entries, for example), not to mention the fact that the formats they impose are often foreign to my experience. The cite journal template, for example, boldfaces volume numbers for journals, which is something I see mainly in scientific journals, almost never in the humanities. This can be overcome with guile, of course, but isn't that contrary to the purpose of templates?—Jerome Kohl (talk) 23:11, 18 January 2013 (UTC)

My comment was in response to what you said: "Why not cut out the middleman and use the sfn template the way it was evidently designed to be used—with direct links from parenthetical referencing?" which is a misunderstanding of the way the sfn template is meant to be used. The sfn template works the same as the harvnb template, exept it has the <ref> tag built right into the template code.

Hand-drawn citations are inaccessible to bots such as Citation Bot. Have a look at this edit, where calling Citation Bot resulted in the addition of digital object identifiers to many of the citations, increasing their value and verifiability. It was even able to locate the title for a journal article that had this information missing from the citation. Citation Bot will not be able to perform these improvements for you unless your sources are wrapped in citation templates. You definitely don't want to try to trick the cite templates into not bolding the volume numbers, as Citation Bot will no longer be able to access the material.

There's a script available at User:Ucucha/HarvErrors that can check for you that each Harvard citation (whether using sfn or harvnb templates) actually points to a book that is present in the bibliography, and that each book in the bibliography is actually used as a reference. This adds even further to the verifiability and accuracy of the citations. Regarding accommodating series of books using the cite book template, this is easily accomodated using the | series = parameter. So there's lots of advantages to using citation templates, and in my opinion the {{sfn}} system is the best one to use for articles that are mostly drawing on books and journal articles as their sources. -- Dianna (talk) 00:34, 19 January 2013 (UTC)

Wow. This all sounds even more impressive than GFHandel's original proposal, for which reason I understand even less than I did before. I do feel constrained, however, to point out that I have already confessed my misconstrual of sfn as harv. In that light, and supposing that my proposal to remove footnotes altogether and replace them with the more modern parenthetical referencing format is the most viable option here, how do these refinements apply? (I can only suppose that the | series = parameter business will eventually sink through my dense brain, but at the moment it still looks like a refugee from the 1970s computer-programming languages I abandoned several decades ago in favour of good old-fashionend English. Sorry.){—Jerome Kohl (talk) 06:31, 19 January 2013 (UTC)

My proposal is identical to GFHandel's initial proposal, so sorry if I am not making that clear. Both sfn templates and harvnb templates create parenthetical Harvard citations. The sfn template is superior because it does not need the additional ref tags and the citations collate automatically. Far from being a throwback to the early days of computing, these innovations are the simplest and best way to create citations for this website at present. -- Dianna (talk) 06:45, 19 January 2013 (UTC)

Well, you leave me in the dust, I am afraid. So long as those ugly footnotes are removed and replaced with proper up-to-date parenthetical references, I bow to the superior technology. I just don't see how that template can be used in this way.—Jerome Kohl (talk) 06:51, 19 January 2013 (UTC)

Which template? if you mean the | series = thing, that's not a template, it's a parameter from the {{cite book}} template. You were wondering how to incorporate the name of a book series into a cite book template, quite a ways up the page. -- Dianna (talk) 06:55, 19 January 2013 (UTC)

I meant the {{sfn}} template, which I believe is unsuitable for Harvard referencing. My concerns here, by the way, are not that the current referencing style can or should be improved. I think I have already made plain that I cannot imagine any change to be an improvement. My worry is rather whether the format imposed by this particular set of templates is really the most desirable one for this article. If you can convince me that it is, or that the advantages of the templates themselves override any disadvantages inherent in the format they use, then we should have no problem achieving consensus here. This has not happened yet, but I am still listening.—Jerome Kohl (talk) 19:58, 19 January 2013 (UTC)