Wikipedia talk:Glossing sources

Request for comment: proposal
After a relatively short exchange at the Village Pump, where I got some good feedback from a few editors, I've decided to throw this open as a general proposal for a guideline ( not a policy ).

In my experience, particularly on humanities-related articles, accompanying a reference with a quote has been very helpful in collaborating over sources. In particular, it helps make sure that over time references don't become associated with statements they don't support. It also helps to prevent statements supported by a reference from gradually changing in meaning, while the reference stays in place. I also feel it helps to raise the level of confidence among editors in sources that are cited. These are concerns specific to the way we write articles on Wikipedia.

From the discussion at the Village Pump, it's clear to me that quotations are not always possible or desirable (for numerous reasons). So the proposal here takes in concerns raised there. I expect more suggestions will come out of this RfC.

I want to emphasise that I intend this as a guideline only - and a loose one at that. I do not think it should be obligatory (repeat: not obligatory ). If others don't want to give a gloss with references they add (or every reference they add) then I don't think that is such a big deal. But I do think it can be good practice. So the proposal is more a matter of "good practice" than anything else.

If you take a look at the example on the proposal page, it might make it clearer what I mean by the advantage of adding a gloss of one kind or another to references. V I've put a notice of this at the Village Pump and Cent. Any ideas for further places to advertise?

--RA (talk) 22:04, 3 October 2010 (UTC)
 * I think that a guideline would be helpful in clarifying the use of this, but basically, it boils down the writer's common sense in where to actually use a quote from a source. / ƒETCH COMMS  /  22:56, 3 October 2010 (UTC)
 * I agree that we should encourage the use of short quotes within references. However: 1. Don't call them "glosses". Wikipedia has enough impenetrable jargon already. "Quote" is far preferable to the obscure scholarly term "gloss". 2. The encouragement of quoting sources in references shouldn't be a separate guideline, it should be added succinctly to Citing sources. This proposal goes into much more detail and at far greater length than is needed. There needs to be advice on how to format the quotes. Several of the cite templates already include a "quote" parameter. There's an issue of what to do when a source is used to support more than one point in an article: how do you format/arrange the quotes in that case? Fences  &amp;  Windows  23:22, 3 October 2010 (UTC)
 * I agree about "gloss" but couldn't think of a better word - the issue being what to call it when it is not a quote e.g. when it is a quick summary in invisible text. Very open to suggestions for a better word tho. I'm tired of jargon too.
 * Agree also about adding something succinctly to Citing sources but a guideline page elsewhere that goes into detail surely wouldn't hurt?
 * About the use of the same ref in multiple places, I thought something like  and then elsewhere  . While that is not of much use to readers, it is of use to editors. It is a similar issue to that faced by how to give different page numbers when re-using refs. --RA (talk) 23:49, 3 October 2010 (UTC)


 * To be honest, I strongly disagree with this proposal. I believe that, in most cases, it won't "glossing" is actually bad practice, not a good one. If we're doing our job right, a large portion of our information should not be so directly sourceable to specific, identifiable quotes.  In the examples you gave in the Village Pump, the info easily and directly corresponds to a specific part of the source text.  But very often, a single sentence in Wikipedia summarizes a paragraph or more of information in a source.  I neither expect nor want that source quoted.  Furthermore, even short quotes on an article with 20 sources will make the references section overly long; if you take a major article, like say Barrack Obama with 273 sources, you'd have a Reference section that would overwhelm the main article.  To be even more honest, I would actually say that we should almost never quote sources in references.  We're an encyclopedia, not an academic journal.  While you may, in some circumstances, be correct in saying that glossing in references makes collaborative editing easier, it definitely makes reading articles much harder.  Qwyrxian (talk) 23:34, 3 October 2010 (UTC)
 * A brief summary would be better than extended quotes for the reasons you give. Use of (so it is hidden from readers) or  (so the enclosed enclosed text isn't even downloaded for readers) or using a dud parameter on cite templates would prevent a lengthy references section/slow download time. --RA (talk) 00:03, 4 October 2010 (UTC)
 * The suggestion to use a comment field or a no-render seems like a step in the right direction, because at least it solves the problem for readers. I still worry, though, on a long article--what happens if an editor with a poor quality computer wants to edit a long article with massive commented/non-rendered commentary?  I'd also say that it would only work well if we enforced list-defined references.  If quotations were embedded in in-line citations, it would be nearly impossible to edit, as you'd have to look really hard to figure out what is actually part of the article and what is part of a quotation that the article is based on.Qwyrxian (talk) 00:12, 4 October 2010 (UTC)
 * Using this proposal only with list-defined references would seriously degrade its utility in terms of matching the reference content to the material it supports, because you couldn't immediately see the connection between the two without a distinct effort to do so. — Gavia immer (talk) 00:51, 4 October 2010 (UTC)
 * You seem to be overlooking Infobbox data and similarly simple information; quoting can be very helpful to ensuring accuracy in such cases. --Cyber cobra (talk) 02:18, 4 October 2010 (UTC)
 * Badger alert! "Quoting" the data in the infobox of, say, Andrei Pervozvanny class battleship, requires copying of three source pages, heavily loaded with tables. Would you accept scanned pages or do I have to actually recreate the original typography? Oh, forgot, it's not in English - should it be rendered in English as well? At a closer look, the very first line of numbers ("Built: 1904–1911") has no underlying quote: the corresponding article section takes up three computer screens, the corresponding sections of two books amount to about fifty pages. Should I quote all fifty or just the first and the last page? End of badger alert. East of Borschov 02:37, 4 October 2010 (UTC)
 * What do badgers have to do with anything? Anyhow, your example seems like an edge-case. Most country articles' Infoboxes, for example, would be quite amenable to "glossing". Certainly choosing whether to gloss requires case-by-case judgment though. --Cyber cobra (talk) 05:24, 4 October 2010 (UTC)


 * A guideline that's not obligatory may rest in peace. There's a practical issue: does every FA candidate need "glosses" for every footnote? Does every "old" FA need revamping to the new standard, or be delisted? No? Then there's no guideline. If it's not required for FAs, it's not necessary at all. East of Borschov 02:27, 4 October 2010 (UTC)
 * I occasionally add a quote when I think the referenced statement might be challenged, or changed without the reference being checked. Less often I'll add a short explanation, as I've seen in some books, if I think it would be helpful to the reader. That said, I think that making this a guideline might be too much - a simple "suggestion" that one can do so when appropriate would be better—since an article that had most or all footnotes glossed would be overkill. First Light (talk) 04:34, 4 October 2010 (UTC)

And what if (as with many, many statements in Wikipedia) the source isn't amenable to trivial quotations? What if you're trying to condense the major point behind a long chapter, or even an entire book, into a single sentence or a short paragraph? That type of source is equally susceptible to this type of "editing accident". In short, this is probably a solution in search of a problem, and even if implemented, would only solve a small number of the simplest cases. I have no objection to collecting 'best practices' and 'advice to editors' in an essay, but I oppose making this page a guideline, and I think that a more practical recommendation is to tell editors to be careful when copyediting without having read the sources. WhatamIdoing (talk) 05:29, 4 October 2010 (UTC)
 * I think there is an issue here that needs to examined, even if the proposed solution isn't practical. The problem is that the position of the footnote gives only a vague indication of what material is being sourced. If you put a footnote at the end of sentence, is it supposed to support the entire sentence or just the last clause? In more complicated sentences, determining the scope of what is being referenced can be more tricky. For example if source A gives the distance and diameter of the moon the an editor might phrase the information as: "At 250,000km from the Earth, the Moon is 3,476km wide at its equator.[A]" Another editor adds further information from a different source: "At 250,000km from the Earth, the Moon, which is not made of cheese,[B] is 3,476km wide at its equator.[A]" Now it appears that the first clause in the sentence is unreferenced though it's actually still supported by source A. The situation gets worse in scientific referencing where a footnote may indicate the source for a paragraph or a section. In the past I've tried to clarify a bit by putting in footnotes like "Jones, p. 345 for the section unless noted otherwise," or "Follows Smith p. 678" to indicate that more than a single statement is supported. It may seem unlikely that situations like these would appear often but from my recent experience working on a cleanup project, Wikipedia talk:Requests for comment/Jagged 85/Cleanup, where it came into question whether the sources support the facts, I'd say it happens more often that you would think. Working on this is difficult because with rewrites, overwrites and deletions taking place over a period of years it's very difficult to determine which facts were added by which editor and which source was intended to support these facts. I think it's an ongoing issue that the task of verifying sources can be complex and can take nearly as much time as gathering the information in the first place. Adding to this the fact that it's impossible to tell whether material has already been checked or how recently if it has, and that fact checking is a task that few editors are willing to do, it's easy to see why so many articles are poorly referenced and errors creep in. The current model of referencing was intended for published work where facts would be checked once by a person dedicated to the job and the author is presumed to be unbiased and have a certain level of expertise. In WP, as we have all seen, incorrect material is often added even with the best of intentions, so repeated checking of sources, with no presumption that the author is unbiased or expert, is needed. Some creativity is going to be needed to develop a new model of referencing that is optimal for Wikipedia, The proposed solution is a start, though it may not be the best, but I think the status quo is becoming increasingly impractical and some changes are needed.--RDBury (talk) 05:02, 4 October 2010 (UTC)
 * There isn't anything preventing editors from doing this today if they find it valuable in certain instances. I'm opposed to adding more work for editors that add sources, there is plenty already. The glosses will surely become old as well when someone references a different page out of the same book down the road. I think it's better to encourage editors to occasionally double-check the sources for an article as part of a cleanup pass. Whenever I do this on a long-running article like soybean, I find that about 75% of the sources have problems and a guideline on glosses wouldn't have prevented most of the decay. —UncleDouggie (talk) 05:14, 4 October 2010 (UTC)
 * The example given can be easily sorted by looking at an article's history. Furthermore, it seems a little silly to write "The diameter of the moon is X.  Um, isn't that what I just said, right there in the sentence?  Isn't repeating it in the citation redundant, and likely to lead to pages that are 50% citations?  (Just imagine Schizophrenia with only one short sentence times each of the more than two hundred separate sources.)
 * I'm absolutely, unalterably opposed to any provision requiring the insertion of direct quotations from the source into footnotes. An enormous amount of extra typing and jamming up of the edit section with no added value. Not to mention it's hard enough teaching noobs how to footnote already without making the process even more complicated. Also, I note that new "not obligatory" directives have a way of becoming obligatory down the road... —Carrite, Oct. 4, 2010.
 * I think all that's really needed is a line or two in WP:Citing sources, and/or the various citation template docs, describing how and when to use the  parameter. --  &oelig; &trade; 16:47, 5 October 2010 (UTC)
 * That's right. First Light (talk) 16:58, 5 October 2010 (UTC)
 * The suggestion that an additional explanatory note (I'm fine with calling it a "gloss," because I am familiar with that term from my higher education, but agree that this would be mysterious at first to many new editors) would be very helpful for several dozen articles on my watchlist. As an optional practice, applied by sound editorial judgment, this could help immensely with verifiability (and thus with neutrality of point of view) in many articles on contentious subjects. Anything that makes sourcing better tends to make Wikipedia better. -- WeijiBaikeBianji (talk, how I edit) 23:07, 29 October 2010 (UTC)

Dialog approach
OK - it looks this isn't going to fly for the good reasons stated by those in opposition. One concern I had when proposing this is that it is a pre-wrapped "solution" rather than something emerging organically from current widespread practice. The question about whether an FA should have "glosses" is a good one: the obvious answer is 'no' ... because this is not current widespread practice.

However, I think there are at least enough who can see the problem, whether or not this (as it is presented here, or anything even like this) is the solution. I'm looking back up at Fences and Widows suggestion to put it in Citing sources and WhatamIdoing's 'no objection' to an essay or guide to best practice. Maybe an approach might be to downgrade this proposal to an essay, or leave as a semi-permanent draft for an extended time. Someone may also be interested in creating a similar page for best practice in (copy) editing vis-a-vis maintaining references. Someone else might want to prepare a best practice on where to place references so that it is clear what they refer to. The point of all this being to initiate a dialog among approaches that may - in time - develop into a widely accepted (and probably multi-stranded) good practice. A section in Citing sources that mentions the problem could link to these essays/permanent drafts as part of encouraging that dialog and getting the wider community to think about the issue.

--RA (talk) 10:09, 4 October 2010 (UTC)
 * I think the issue is, there's no "best way" to determine "is a quote necessary here?", so an essay might be useful in describing this process, but a formal guideline won't ever actually be completely binding. Yes, this is useful sometimes, and no it is not appropriate always, so it might be good to detail some examples where it would work and where it would not, but any actual guideline is defeated by the "it's not required" part. / ƒETCH COMMS  /  12:49, 4 October 2010 (UTC)
 * The best guidelines aren't obligatory - they guide :) I think some expansion on why, when, and how to give quote from sources in references to support text would be a good idea. An essay might be a good idea, and some more mention of the issue could go in Citing sources. I've done it when the fact seemed somewhat unlikely or controversial, or when the text was not easily accessible. Fences  &amp;  Windows  22:17, 4 October 2010 (UTC)
 * I'd suggest that the need for glosses varies according to the sources and the assertions. Obscure sources benefit by having the cited material quoted. So do remarkable claims, or assertions whose character is dependent on the precise (and perhaps ambiguous) wording of the source. For those reasons, the diameter of the Earth examples may not be the best recommendations for this proposal.   Will Beback    talk    00:32, 5 October 2010 (UTC)
 * I agree with the comments above that it would be helpful to have a mention at Citing sources—something like "In addition, a short quote from the referenced source can be added to the footnote if an editor feels the material might be perceived as unlikely or controversial, or if readers would be helped by having additional explanation" (that is off the top of my head, and only an example of what might be said there). First Light (talk) 04:50, 5 October 2010 (UTC)

Essay-fied
Thanks everyone for your input. I've essay-fied it. --RA (talk) 21:31, 7 October 2010 (UTC)