Wikipedia talk:Citing sources/Archive 36

I wonder if we might consider whether it is useful to to distinguish "in-text" citation (defined at WP:Citing sources#In-text attribution as an attribution within a sentence, "in addition to an inline citation after the sentence") from "inline citation". Quite aside from the curious suggestion of "in addition to", this distinction between "in-text" and "inline" seems to be made only here. The earliest use of this term (at WT:Citing sources/Archive 1#Proposed in-text citation guidelines in 2004) did not make this distinction, nor have I found any other discussion of the distinction. Note that I am not suggesting revisiting whether "mid-sentence footnotes" should be deprecated (as discussed here); I ask if we really need a special term (and an inapt one at that) just to deprecate it. I am wondering if that section could be merged with the "inline citation" section. ~ J. Johnson (JJ) (talk) 23:26, 4 June 2013 (UTC)

In-text attribution is "Smith argues that ..." It's not the same as inline citation – the distinction disappears when you're using Harvard refs, but most editors don't. Using in-text attribution is important to avoid allegations of plagiarism, so I wouldn't want to see it changed here. SlimVirgin ^(talk) 23:45, 4 June 2013 (UTC)

Huh? Well, "Smith" is incomplete, but presuming there is a year or something to make that an adequate short cite: it is a (short) citation, and it is "located near the material...", so how is it not an inline citation? How does this alleged distinction disappear with "Harvard refs"? ~ J. Johnson (JJ) (talk) 00:10, 5 June 2013 (UTC)

The most popular style I know of that uses notes for inline citations is the Chicago Manual of Style. The 16th edition in section 4.98 states "Credit lines. Whether or not the use of others' material requires permission, an author should give the exact source of such material: in a note or internal reference in the text, in a source note to a table, or in a credit line under an illustration." I take this to mean a note number next to a quotation and a corresponding note which is a short or full citation is sufficient. Naming the author within the running text is not required by Chicago, as I understand it. If so, this guideline contains a Wikipedia house preference, and not a requirement that is widespread in the English-speaking world. Jc3s5h (talk) 00:50, 5 June 2013 (UTC)

Smith is not incomplete. Smith may not be the source; he may be someone interviewed in the source. In-text attribution is not an inline citation. The two can blur in the case of Harvard refs because where Smith is the source, you might write Smith (2013), in which case you have in-text attribution via a Harvard ref (via an inline citation). SlimVirgin ^(talk) 00:14, 5 June 2013 (UTC)

I believe the key here is "someone interviewed in the source". That is: "according to Jones (2004) Smith said ...", which is indirect attribution. Smith is the original source, but "source" in the sense of "work consulted" is "Jones 2004". Compare this to direct attribution: "Smith (2002) said ...."

That both the named section and the definition (in "Types of citation") completely miss this distinction, and the former makes the totally misleading distinction that in-text attribution is "inside a sentence" while inline citation is "after the sentence", is a prime demonstration of just how FUBAR-ed (any one need that spelled out?) things have gotten here. This stuff needs a total rewrite. ~ J. Johnson (JJ) (talk) 19:38, 5 June 2013 (UTC)

I believe the case where in-text attribution is important is to emphasize that a certain view belongs to the person it is attributed to, and not necessarily the view of the community of Wikipedia editors. If the statement merely has quotation marks and a footnote, it might interpreted as a statement of the mainstream view, quoted because the person who wrote it came up with a particularly apt statement of the mainstream view. On the other hand, it was wise for whoever wrote our article on Richard Nixon to use in-text attribution for the sentence "I'm not a crook." [Kilpatrick, Carroll (November 18, 1973). "Nixon tells editors, 'I'm not a crook'". The Washington Post. Retrieved July 17, 2011.]

I've never understood the distinction between in-text (used by Chicago and others) and inline (used only on Wikipedia). But I ignore it and things work just fine. --  Gadget850 ^talk 02:32, 5 June 2013 (UTC)

Your lack of understanding is because the claimed distinction exists (if it exists at all) in minds of certain editors, not in any plain text the rest of us can share. ~ J. Johnson (JJ) (talk) 19:41, 5 June 2013 (UTC)

It's not complicated:

Richard Nixon said, "I'm not a crook."^[1]

The bit that runs "Richard Nixon said" is WP:INTEXT. The little blue number (and the bibliography entry that it leads to) is WP:INLINE. WhatamIdoing (talk) 23:05, 5 June 2013 (UTC)

And it's not a phrase used only on Wikipedia. See here, for example (bold added):

All ideas, facts, figures and quotations taken from external sources must be properly cited. ... In addition, all direct quotations require an in-text attribution. ... For example, you cannot simply insert the quote “All is flux” into the body of your text. You must include a phrase such as 'According to Heraclitus or “As Heraclitus stated” or “In the words of Heraclitus,” “All is flux.”

SlimVirgin ^(talk) 23:15, 5 June 2013 (UTC)

SV Just as well talk pages do not have to meet the standards laid out in the guideline otherwise you ought to have included MCTC Philosophy Department :-O

The requirement in the content policy has been for may years that a POV must have inline/in text attribution, the requirement for all quotes to have in-text attribution has never been agreed. I think that there are lots of example where in-text attribution is not needed for example with well known phrases such as "second to none". One might summarise several sources and write a sentence They were "second to none".^{[1 (bundled)]} or as an alternative They were second to none.^{[1 (bundled)]} Whether one puts "second to none" in quotations is largely a matter of style (there is no copyright requirement). What one does not have to do is alter the sentence with quotation marks to They were, in the words of the well known traditional phrase, "second to none".^{[1 (bundled)]} to meet any Wikipedia content policy. Like so much else it is convenient to for editors to use separate terms for these concepts and a I for one am comfortable with using "in-text" and "in-line" (with or without hyphens) to describe the difference. The only place were this becomes somewhat confusing is when "Parenthetical referencing (Harvard style)" in-line citations are used, but I think that the examples in the guideline make the distinction clear. -- PBS (talk) 08:16, 6 June 2013 (UTC)

• There is consensus that quotes need in-text attribution so far as I know. For example, see WP:MOSQUOTE: "The author of a quote of a full sentence or more should be named; this is done in the main text and not in a footnote." If you don't attribute a quote in-text, it may look like a scare quote. SlimVirgin ^(talk) 22:13, 6 June 2013 (UTC)

Sure, not complicated at all. "In-text attribution" is defined in the section "Types of citation", which certainly implies it is a type of citation. And this is reinforced by the statement of when it should be done (quotations, etc.), which parallel the cases that require citation. What, then is the difference between this and "inline citation"? We link to the corresponding section, and the gist of the first sentence is that "in-text attribution" is inside a sentence, while "inline citation" comes after the sentence. So the example WhatamIdoing provided just above is inline citation, while the following example is in-text attribution (which is just a form of citation) because it is inside the sentence:

Richard Nixon said,^[1] "I'm not a crook."

Yes, quite clear. In fact, I think some of you have a different distinction in mind (and possiby a valid one at that), but you have failed to reduce it to plain text. ~ J. Johnson (JJ) (talk) 22:40, 6 June 2013 (UTC)

In the INTEXT attribution points to a publication, then it is also, simultaneously, an INLINE citation. The example of Richard Nixon is not an example of both (Nixon is not a publication), but "According to Alice Expert in her 2010 book, The Sun is Really Big..." is both INTEXT and INLINE. WhatamIdoing (talk) 04:01, 8 June 2013 (UTC) Likely you mean "If ..."? -JJ

Sorry, but that is not the plain reading of the text. And your added distinction of pointing to a publication just makes things more complicated without really resolving the fundamental problem, which is confounding attribution with citation. One is the means by how the other is done. ~ J. Johnson (JJ) (talk) 19:25, 8 June 2013 (UTC)

P.S. Regarding the examples, you may have confused to whom (Nixon/Expert) a statement is attributed with the citation of where (the publication) the statement is documented. ~ J. Johnson (JJ) (talk) 19:34, 8 June 2013 (UTC)

In this section a discussion is made which considers whether a single Wikipedia article should contain references using multiple citation formats. Blue Rasberry (talk) 14:20, 30 July 2013 (UTC)

This edit request has been answered. Set the |answered= parameter to no to reactivate your request.

the player is now without a club.sources sky sports who report than he has not been offered deal by morecambe Mandarj123 (talk) 15:16, 31 July 2013 (UTC)

What? -- JHunterJ (talk) 15:23, 31 July 2013 (UTC)

Not done: this is the talk page for discussing improvements to the page Wikipedia:Citing sources/Archive 35. Please make your request at the talk page for the article concerned. --Redrose64 (talk) 16:08, 31 July 2013 (UTC)

How do I properly cite different sections of a book whose pages are not numbered? For instance, this book. In Doug Stone, I want to cite the chapters on both him and Doug Johnson. Would I just put something like "Miller, Zell (1996). 'Doug Stone'"? Ten Pound Hammer • ^{(What did I screw up now?)} 21:43, 31 July 2013 (UTC)

That book does have page numbers, at least on most of the pages. To find the page number of a page that is not numbered, just look a the page before or after the page in question. Jc3s5h (talk) 21:55, 31 July 2013 (UTC)

Or look at the handy "Page" counter Google supplies at the top of the preview. Note that this could be out of sync with the actual printed page numbers so it is prudent to check on a page that does have a printed page number. But in this case it matches. ~ J. Johnson (JJ) (talk) 22:03, 31 July 2013 (UTC)

In a continuing effort to clarify this guideline by removing confusing and meaningless verbiage, I propose replacing the first two paragraphs with the following:

Citation (broadly) is the means of connecting material in an article to a source. Citation is the basis of verifiability, a core content policy. As stated in that policy, all material in Wikipedia articles must be verifiable, and certain kinds of material must be explicitly cited to a source. Other policies set requirements regarding appropriate sources and their use (see no original research, reliable sources, neutral point of view, and copyright policy); these will not be discussed here. This guideline is intended to provide guidance in the practice of citation on Wikipedia.

I point out that this is only a start to the guideline, and neither does nor should cover every possible point, exception, or misinterpretation in the first paragraph. Likewise regarding detailed definitions and uses of "citation as a thing", general references, etc.: no stand on these is taken here, I am just trying to create some basis on which we can stand. ~ J. Johnson (JJ) (talk) 22:00, 26 May 2013 (UTC)

late interjected comment At first blush, I think that the proposed paragraph is allaround better in that it deals with the generalities and scope rather than directly diving into the specifics of implementation. (I'm starting a review of this conversation, so I thought I'd provide a first impression of the initial post). --User:Ceyockey (talk to me) 00:02, 24 July 2013 (UTC)

The proposal essentially says a general reference is not a citation, but it is, so the proposal fails. Jc3s5h (talk) 22:31, 26 May 2013 (UTC)

I also oppose this proposal, and invite the proposer to drop the stick and stop beating this dead horse. The act of citing sources is meant to show your work, which may or may not involve connecting specific material to specific sources. Citation may be an action, but a citation (what we're defining in the first sentence) is a thing, not an action. WhatamIdoing (talk) 23:23, 26 May 2013 (UTC)

Jc: the proposal does not "essentially say" anything about general references. It is as I just said regarding "general references, etc.: no stand on these is taken here." (Emphasis in the original.) Nor does it say anything about what a citation is, or is not. Your objection is effectively pulled out of thin air, as it has no basis in the text of the proposal.

W: you always oppose any change to improve Wikipedia, so what's new? Perhaps you wouldn't mind getting out of the road? It also does not help that you don't pay attention. (How many times do I have to ask you?) E.g., you say that "what we're defining in the first sentence..." — ah, is that in the first sentence of the guideline, or first sentence of the proposal? What you seem to have missed is where I said the proposal is to replace that "first sentence". And the proposed replacement explicitly states it is referring to citation broadly, not "a" citation. We can define your "line of text" kind of citation at a lower level. There is room for that, and "general reference", within the broad scope I am trying to sketch, but it is probably a waste of time showing you how, as you are such a blind obstructionist. ~ J. Johnson (JJ) (talk) 22:50, 27 May 2013 (UTC)

A citation in the terms of WP:V can be read to mean inline citation. There may be less objection if the first sentence was "A reference is the means of connecting material in an article to a source". -- PBS (talk) 08:38, 28 May 2013 (UTC)

interjected late out of time sequence I think I'll stop reviewing there as it just went ad hominem. However, on the matter of general references, I think the proposed paragraph is neutral on those, by no means saying "a general reference is not a citation". With respect to the semantics of citing/citation/citations ... connection of a wikipedia article or material therein to a specific source is an implicit requirement of verifiability; if you can't point at a specific source, then how is it verifiable? It is not. --User:Ceyockey (talk to me) 00:09, 24 July 2013 (UTC)

Can be read, yes, but that's there, and why should that merely possible and non-authoritative interpretation be controlling on an explicit definition here? And again yes, undoubtedly there would be less opposition if in the very first sentence of this guideline we confirm various articles of faith regarding "a citation" and "general references". But that is part of the problem with this guideline — it starts by confirming various bones that certain people don't want to let go of, and then it tries to stitch these disparate parts together. Hell and damnation, WE CAN'T EVEN SENSIBLY DISCUSS various points, because when I say "citation (broadly)" someone else insists it must be "a citation", and so on. In this proposal I am not taking a stand on any of these pet concerns, I am, as I said at the outset, trying to give us a basis on which to stand. Which leaves plenty of room to argue all those points further on. ~ J. Johnson (JJ) (talk) 00:41, 29 May 2013 (UTC)

PBS, it would be far more accurate to say "A reference or a citation is the stuff you type to tell the reader what sources you used to write the article". It doesn't matter which word you're using: the community (for better or worse) has defined reference and citation as being synonyms. The community also defines both references and citations as having two subtypes: the inline subtype (highly desirable) and the non-inline subtype (permitted). Both of these subtypes are still these things, but J Johnson's definition excludes one of the subtypes.

J Johnson, if you define a citation as "connecting material in an article to a source", and general references are correctly defined as "not 'connecting material in an article to a source'" then, yes, you are defining general references as not being citations. WhatamIdoing (talk) 01:51, 29 May 2013 (UTC)

Agreed. The proposed sentence, "Citation (broadly) is the means of connecting material in an article to a source" describes just one use of a citation. A citation, or reference, is simply text that uniquely identifies a piece of work. See, for example, here.

I think the current first sentence is fine as it is; personally, I prefer it with "a line of text," but without is okay too. SlimVirgin ^(talk) 02:01, 29 May 2013 (UTC)

W: So you restrict all references/citations narrowly, assert a "community definition" that there two subtypes, then boldly leap across a logical lacuna to declare that my definition excludes one of the subtypes. Please, can you tell me which of those fourteen words mentions, or even suggests, anything in the nature of a subtype? Or even hints at any kind exclusion?

You also blame me for making "general reference" not a citation, because I define citation as "connecting material to a source", and you define "general reference" as not connecting such material. No, that is not consequence of my definition, that is your restriction. I suspect you are thinking that genrefs don't connect from the text (i.e., "inline"), but (ha ha, fooled you) my definition does not specify that. My definition explicitly says broadly; it says nothing at this level as to the nature of the connection, or the means. I do not exclude genrefs, you do. I point out that the definition at Citation also excludes genrefs, whereas my approach is permissive.

I point out to all that the narrow usage of "citation" as a descriptive line of text — i.e., the bibliographic reference, or, as I call it, the full citation — is subsumed as part of the means by which "citation broadly" is done. Citation narrowly gets described at a lower level of detail. My proposal starts by describing the whole puzzle, not just one of the pieces. ~ J. Johnson (JJ) (talk) 22:08, 29 May 2013 (UTC)

We're telling you what your proposed text actually says, and it's no good telling us that you meant something else. We've got enough problems in our policies and guidelines with text that doesn't actually mean exactly what a plain reading of it indicates, and I'm convinced that nobody else here is interested in adding any more examples of this problem. WhatamIdoing (talk) 22:40, 29 May 2013 (UTC)

I submit that plain reading of text means without adding your own definitions and interpretations. You are the one that is trying to make fourteen plain words mean something else. ~ J. Johnson (JJ) (talk) 00:03, 30 May 2013 (UTC)

As there has been no comment on the rest of the proposed change I presume it is acceptable. It is in the nature of a scope statement that delineates what is, and is not covered, and the benefit of referencing other relevant policies without trying to redefine them here seems obvious.

As to the initial definition of "citation (broadly)" I reiterate that it is a broad, general statement, which makes no statements nor implications regarding any details of form or usage. (It's like a large block of rock, which has yet to be carved.) As to "general references": while I feel this notion has gotten very entangled, that is largely for lack of proper definition. The proposed definition of "citation (broadly)" provides a basis for properly defining "general reference" and possibly resolving some of those issues, and is actually more permissive of that concept than other definitions. I emphatically state: it does NOT exclude "general reference", nor am I trying to take away anyone's favorite bone. If that is understood, I would like to get on with some needed improvements. ~ J. Johnson (JJ) (talk) 20:19, 1 June 2013 (UTC)

There being no further discussion of the proposal, I have made the change. Two points: 1) There are some places in the rest of the text slightly discordant with this, but we can work those out. 2) It occurred to me that the sentence starting "Other policies set requirements ..." perhaps should be strengthened by prefacing it with "Editors should be aware of ...". Comments? ~ J. Johnson (JJ) (talk) 21:55, 3 June 2013 (UTC)

I restored the first paragraph, which seems clearer. It's important to say that a citation is the same thing as a reference, and to give an example. SlimVirgin ^(talk) 23:56, 4 June 2013 (UTC)

Your after the act remark ("seems clearer") does not arise to the level of discussion, so I have reverted your "restoration". I agree that examples are important, but that comes in at a lower level. If you have any other objections, concerns, or confusion, please raise them here. ~ J. Johnson (JJ) (talk)

Your proposals have been opposed by several editors, because they're less clear than the current version, and arguably introduce an unusual understanding of "citation". Therefore, please either let this go or continue the discussion, but don't try to force in your changes. SlimVirgin ^(talk) 20:14, 5 June 2013 (UTC)

Okay, let's continue. You claim that I 'introduce an unusual understanding of "citation".' I presume you refer to this concept of "connecting" material to a source. In fact that was part of CharlesGillingham's suggestion for a new lead back in 2008. (And excised only because someone complained of verbosity, not for any complaint re the concept.) Possibly other terms could be used ("attribution"?), but "connection" seems to plainly and aptly describe the relationship between "material" and "sources". You have not shown there is anything "unusual" in either the relationship or the term; your complaint is deficient.

Earlier (29 May) you complained that the proposed definition "describes just one use of a citation." Overlooking a tiny error that I did not describe "a" citation, but citation broadly, I ask: what is wrong with that? (Especially as your following definition equally "describes just one use" of the term.)

Please pay attention (W, too): I defined citation broadly, you defined it narrowly (as a citation). Note: these are non-conflicting definitions. This definition of citation-BROADLY does not alter your definition of "a" citation. (Which as I have said before, should be defined at a lower level.) What part of this do you not understand? ~ J. Johnson (JJ) (talk) 00:46, 7 June 2013 (UTC)

No, you said that you were defining the concept broadly, but you then actually presented a very narrow definition. Mere assertion that the definition is broad does not make it so. WhatamIdoing (talk) 03:57, 8 June 2013 (UTC)

What???! What kind of unstated (secret?) interpretations are you applying to make my plain words "very narrow"? To go back to a previous question (which you have yet to answer) which of the proposed fourteen words in any way narrow, exclude, restrict, straiten, or qualify this definition? Or (if analysis is too difficult) just give me a couple of examples of citation that are excluded by the proposed definition. I throw your own words back at you: Mere assertion that the definition is narrow does not make it so. ~ J. Johnson (JJ) (talk) 20:10, 8 June 2013 (UTC)

W: Still waiting for any examples of "citation" that are excluded by the proposed definition. Put up or shut up. ~ J. Johnson (JJ) (talk) 19:45, 9 June 2013 (UTC)

We've already answered you: your definition—the one that says "the means of connecting material in an article to a source"—is invalid for any citation that does not "connect material" to the source, but instead merely provides information about a source without "connecting material" to that source. For example, go look at the very first example at the top of this guideline. That's "a citation" for Wikipedia's purposes. However, it is (1) not in an article and (2) does not connect any material to the source. WhatamIdoing (talk) 20:16, 9 June 2013 (UTC)

  And I have already rebutted you (above, on 29 May). But since you weren't paying attention, let's go over it again. You apply your own interpretation and qualifications, such as assuming that the proposed definition applies only to material in articles (DOES NOT SAY THAT!), or that "connecting material" limits it to inline citations (DOES NOT SAY THAT!). Then (omighosh!) you find that you (I repeat: you) have excluded your golden calf of general reference. Which you then blame on me. Well, I call bullshit on that.

  Your argument that the "Ritter" example would be excluded because it is not in an article is wholly discreditable. In the first place, the proposed definition does not say only in an article; use in other contexts is not excluded. Second, I think it is generally understood that examples used here and elsewhere are to be taken as they would be used in an article. Third, if you want to argue like that, then the Ritter example is not a citation by the existing definition, which says that citations are used to identify the reliable sources on which an article is based, because no article is based on this example. Similarly, all of the "Alice Expert" examples — such as you have used yourself on this page — are not citations, because "Alice Expert" is not a reliable source. (And if you really want to have a pissing contest, then I claim that her book has multiple editions, and in failing to uniquely identify which one your example again fails the current definition, and therefore is not a citation.)

  In summary: your conclusion that the proposed definition is "very narrow" is due entirely to your misinterpretation, and your twisted argumentation. Your objection is invalid. And I tell you again: the proposed definition does not exclude your "general references". ~ J. Johnson (JJ) (talk) 01:11, 12 June 2013 (UTC)

Your definition begins "Citation (broadly) is the means of connecting material in an article to a source.".

Please consider the Ritter example at the top of the guideline, and tell me what "material in an article" that source is connected to. WhatamIdoing (talk) 17:27, 12 June 2013 (UTC)

The Ritter example connects whatever example it is associated with. Presumably the example is presented to show what a "line of text" kind of full citation looks like, without bothering with other details. In fact it does contain a specification (the "p. 1"), which implies that the connection (identification) is to specific material. And it is not only reasonable but quite likely that a citation of precisely this form could be used in the text (with or without parentheses), or in a footnote in the text, in exactly the same way we might expect to see a short cite used. That this example does not show us what material it is associated with is because it was taken out of context. That this or any other example does not actually connect, or "uniquely identify", a source (let alone a reliable source) is immaterial; the point is to show how the citation would be done. ~ J. Johnson (JJ) (talk) 21:23, 12 June 2013 (UTC)

That this example does not show us what material it is associated with is because there is no such material, nor any article containing the material. It is not any less of a citation as a result. Your definition says that citations are not the text describing the source, but a means of connecting material and sources. In other words, your citation says that if there is no material to be connected (e.g., in a ==Further reading== section) then the text describing the source does not meet the fundamental requirements of a citation. This is why your proposed text is inappropriate for our guideline. WhatamIdoing (talk) 15:23, 13 June 2013 (UTC)

Why do you keep misrepresenting the proposed text? The proposed definition does not say (as you just stated) "citations are not the text describing the source". Nor does it say anything about "if there is no material to be connected...". You keep just making up stuff that has absolutely nothing to with the proposed text. Your examples are applicable only to your definition, which is not the one proposed. ~ J. Johnson (JJ) (talk) 20:32, 13 June 2013 (UTC)

I'm not misrepresenting anything.

The proposed text is: Citation (broadly) is the means of connecting material in an article to a source.

The proposed text therefore defines something that does not involve "connecting material in an article to a source" as not being a citation.

The example at the top of this guideline does not involve "connecting material in an article to a source". Therefore, according to your proposed definition, the Ritter example is not "a citation" as you define it. WhatamIdoing (talk) 23:12, 16 June 2013 (UTC)

  Your conclusion ("therefore ...") is invalid because it depends on your assumption that the definition is limited to only "connected material". Where does the definition say "only" or "limited to"? It does not, so please stop stating that it does.

  The situation is exactly as if I had said "1, 2, and 3 are digits", and then you misinterpret that as "only 1, 2, and 3 are digits", to conclude that I said "4 and 5 are not digits". Which is patently false. The proposal says (essentially) that citation is "something" that "connects material ...". Then you insert "only", to conclude that something not explicitly included in the definition (material not connected) is therefore excluded. That conclusion is both invalid and false.

  You are indeed misrepresenting what the proposed definition says. ~ J. Johnson (JJ) (talk) 22:17, 17 June 2013 (UTC)

If you say "X is 1", then, yes, people will quite reasonably interpret this as meaning "X is not 2, 3, 4, or any number except 1". If, on the other hand, you say "X is 1, among other things", then people would assume that X is multiple things.

Here, you define X (citations) as being one thing ("the means of connecting material in an article to a source"). You did not define it as multiple things (e.g., "Citations are the means of connecting material in an article to a source, among other things"). You might not have intended to write so restrictive a definition, but that's what you actually did in that sentence. You should probably chalk this up to "reasons why I shouldn't proofread my own work" and then either give up or change your proposal so that it doesn't have this restrictive definition. WhatamIdoing (talk) 22:30, 17 June 2013 (UTC)

  The issue is not with proof-reading, but fool-proofing. It doesn't matter how carefully I "word-craft" if you insist on kicking everything to flinders.

  In the present iteration you confuse identity with inclusion. The former is akin to the assignment of a value in arithmetic or programming. It would even hold in set theory if one says "set X is exactly the elements 1, 2, and 3, and no others". However, in my example I did not say that. I said that "1, 2, and 3 are digits". Which is not saying that "1 = digits". (Totally nonsensical.) It is saying that the named items are included in a class (or have the characteristic of being) "digits". It says nothing at all about other members of that class. My example says nothing about "4"; your faulty conclusion derives solely from your inclusion of "only" (or its equivalent, "and no others") into the definition. Is that clear enough for you?

  But your confusion runs even deeper, so let's try another analogy. Suppose we were trying to define "nails". I might say: "Nails (broadly) are a means of connecting two or more pieces of wood together." Whereupon you argue that by this definition the bits of metal in the bins at the hardware store are not nails because they are not actually connecting any wood. Sorry, this definition does not say "actually connecting", it says a means of connecting. That a nail in a bin has not yet been used for its intended purpose makes it no less a nail. Same for citations.

  You still have not shown that the proposed definition is in any way restrictive or narrow other than in your misinterpretations. ~ J. Johnson (JJ) (talk) 19:15, 19 June 2013 (UTC)

In your analogy to nails, the bits of metal at the hardware store might still be nails, but the ones that have been permanently turned into wood-free artwork would cease to be nails under your definition.

I don't think there's any point in continuing this conversation. Writing policies and guidelines is hard precisely because we need to fool-proof definitions as much as possible. If you don't want to work within that constraint, then go write essays. WhatamIdoing (talk) 11:40, 20 June 2013 (UTC)

  Whether metal artwork used to nail together wood would be a "nail" under any definition is a question I leave to the Wikilawyers. In the present case, I say that the metal objects in a bin in a hardware store are still "a means of connecting two or more pieces of wood together", and therefore are nails by the definition. (No "might" about it: are.) Similarly for the proposed definition of "citation (broadly)": a means of connecting, regardless of whether any individual instance actually connects, or not.

  To the particular point you have been trying to argue: although the Ritter example above does not in this case connect any material in an article to a source, it still is a means of doing so, and so is not excluded by the proposed definition. And you have yet to show an example of a citation not included in this definition.

  If you won't work within the constraints of not misinterpreting what has been proposed or said, nor making absurd objections, then I extend the same suggestion to you: please go. If you can point out any real problems, or have any sincere concerns regarding the actual proposal, or suggestions for improvement, or even complete alternatives, I am willing to continue. ~ J. Johnson (JJ) (talk) 22:49, 21 June 2013 (UTC)

Hi, a question about infoboxes. The relevant bit of the policy is If an infobox or table contains text that needs citing, but the box or table cannot incorporate an inline citation, the citation should appear in a caption or other text that discusses the material. If both an infobox and the main article text contain identical content (eg a date of birth), do we need duplicate citations in both the infobox and text? Most articles seem to only have them in the text. Is this explicity stated anywhere? It is possible to have an inline citation in the infobox, but it does spoil the appearance a bit. SmilingFace (talk) 22:08, 5 August 2013 (UTC)

I think it can be treated as part of the lead; i.e, you don't need to give a citation for information that is properly cited elsewhere in the article. Zero^talk 00:01, 6 August 2013 (UTC)

I disagree; the cite is buried way down in the main body of the article, and the fact that it wasn't cited was enough for me to challenge and remove it. What harm is there in citing it directly? GiantSnowman 08:09, 6 August 2013 (UTC)

WP:WHYCITE states

Citations are not used on disambiguation pages (sourcing for the information given there should be done in the target articles). Citations are also often discouraged in the lead section of an article, insofar as it summarizes information for which sources are given later in the article, although such things as quotations and particularly controversial statements should be supported by citations even in the lead.

In my view, it should be easy to figure out what part of an article discusses a statement in the lead, and an article (which GiantSnoman neglected to name) which fails to make that clear needs to have the lead revised. At the same time, I think it is reasonable to expect readers to make some reasonable effort to make the connection; such efforts should include reading the entire article, and searching for key words or dates found in the lead by using the browser's feature for searching within a page. Jc3s5h (talk) 15:58, 6 August 2013 (UTC)

A few things:

1. An infobox is not the same as a lede.
2. I didn't neglect to mention the name of the article, SmilingFace did in their OP - it relates to Tony Anthony (evangelist).
3. Given the controversy/issues relating to this individual's name/background, explicitly citing information is simply common sense, especially from a BLP perspective.

Surely we should be encouraging more references and cites on Wikipedia, especially about living people! GiantSnowman 16:06, 6 August 2013 (UTC)

I see no reason to distinguish between a lead and an infobox in this situation. I have never seen any scholarly publication that made a practice of repeating citations for the same fact everywhere the fact appears. Jc3s5h (talk) 16:25, 6 August 2013 (UTC)

Some basis for that view in policy would be appreciated. GiantSnowman 16:27, 6 August 2013 (UTC)

Wikipedia has an essay, but not a policy about this: Wikipedia:Citation overkill. Also, Wikipedia is edited as a scholarly publication. At least one scholarly style manual, the United Nations Editorial Manual, supports the idea that repeated citation of the same information is not normally required:

Once an item has been referenced, whether in a footnote or text note, the reference is repeated only when necessary for the sake of clarity or to change a specific element in the reference, such as a section or paragraph number.

Jc3s5h (talk) 17:25, 6 August 2013 (UTC)

I have two degrees, I know how to cite in an academic context - which this is not. Academic manuals say you cite first and then any subsequent mentions do not need citing - whereas you are saying cite the last item? GiantSnowman 17:53, 6 August 2013 (UTC)

In an academic context, you do cite the first instance - in the main text. If the same fact is mentioned in the introduction (or in the abstract), that doesn't count as the first instance. The introduction of a paper corresponds to the lead of a Wikipedia article, so, by extension, the first mention in the main text of the Wikipedia article is where the ref should go. --Redrose64 (talk) 19:10, 6 August 2013 (UTC)

But the infobox is neither. Plus something to ponder over - today's featured article has lots of cites in the infobox. 19:20, 6 August 2013 (UTC)

Once per article is accepted as being enough. Whether that "once" should be the first one as determined by vertical height on a particular user's computer screen, or first according to what you see if reading the prose, or first according to the prose but counting only after the first section heading, or even not first, but at the most important mention (e.g., an entire section devoted to it), is really not worth writing an official rule over.

There is and has never been a rule that the same fact must be cited twice on the same page. If a fact appears twice, and you don't notice the citation is at the other, then you may challenge it (after all, you honestly thought it was uncited), but when your error in overlooking the citation is pointed out, your challenge has failed and the material can be de-tagged (if you fact-tagged it) or restored (if you removed it). WhatamIdoing (talk) 22:00, 6 August 2013 (UTC)

Again, I will repeat - what is the harm in directly citing something in the infobox? It's used in FAs for goodnesss sake, and avoids information getting removed or queried in good faith. GiantSnowman 10:02, 7 August 2013 (UTC)

• There is another growing issue with infobox citations. Some editors are simply dropping several citations, using the cite templates) at the end of an infobox. These show up on a new line containing bracked numbers, at the end and there is no indication what elements in the box these are for. So if, lets say, three citations are added, then for any element either 0, 1, 2 or 3 of these could apply to each element. That for me is less then useful. We probably should address this also in this discussion. Vegaswikian (talk) 19:18, 6 August 2013 (UTC)

• You should directly cite the information in the infovox, just as you would directly citein the main body. GiantSnowman 10:02, 7 August 2013 (UTC)

Yes. This seems analogous to the issue of where to cite multiple facts in a single sentence. It seems some editors prefer having all of the links at the end of the sentence (for being neater?), but this loses the connection of which citation goes to which fact. I say have each citation immediately adjacent to the material it supports. ~ J. Johnson (JJ) (talk) 18:44, 8 August 2013 (UTC)

This came up in Talk:Otaku/GA1 and I don't really know what to do - can we allow book without page range? (Please ping me with @[[User:Piotrus]] when you reply - I'd appreciate it. Thanks! --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 08:25, 13 August 2013 (UTC)

WP:V does not require page numbers; lack of page numbering must not debar an otherwise reliable source. --Redrose64 (talk) 09:26, 13 August 2013 (UTC)

Agreed, WP:V does not require it ... but there are ways to make the finding of content easier. Using the _chapter_ parameter is one; using the _at_ parameter another; and including a quotation via the _quote_ parameter is a third. --User:Ceyockey (talk to me) 11:16, 13 August 2013 (UTC)

I do not know how to use the chapter parameter for the templates, but I should learn it. It concerns two sources. @User:Piotrus (as requested above) ChrisGualtieri (talk) 12:59, 13 August 2013 (UTC)

Sounds like a good idea, thanks for the help. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 13:07, 13 August 2013 (UTC)

You might put |chapter=Chapter 6 or if there is a longer title, |chapter=Chapter Six, in which Eeyore has a birthday and gets two presents --Redrose64 (talk) 14:16, 13 August 2013 (UTC)

There's a discussion at Help talk:Citation Style 1, about whether it's necessary to note the medium through which a source was accessed (eg. Google Books), which could benefit from some input by people more familiar with the guidelines. It's really more of a series of discussions – start at this section and scroll down. DoctorKubla (talk) 06:42, 12 August 2013 (UTC)

This appears to be the perennial claim that "SAYWHEREYOUGOTIT" requires people to say that they "got it" from their library, from Google Books, etc. We should probably put an explicit denial of that idea back into the guideline. WhatamIdoing (talk) 21:21, 14 August 2013 (UTC)

Hi. I didn't find anything in the archives or the FAQ so maybe some one can clarify this for me: Even if the Wikipedia:Manual of Style request the use of endashes in some cases or says that a straight apostrophe should be used instead of a curly apostrophe, should I change the title of a reference in order to adapt it to the typographic conventions? Even if this means that spaces or quotation marks are removed? --Jaellee (talk) 13:03, 15 August 2013 (UTC)

Apostrophes are always straight, that is just an appearance item, similar to different ways the ampersand might appear. But the variety of dash involves shades of meaning; if the person who created the title of a work decided that an en dash or a hyphen best reflected the meaning, I would leave it alone. Jc3s5h (talk) 14:48, 15 August 2013 (UTC)

Replace

...the dead URL is not necessary. Simply remove it.

with

...the dead URL is not necessary. Simply remove the dead URL leaving the remainder of the reference intact.

after recently finding several people deleting the whole reference. -- 86.128.54.183 (talk) 19:18, 18 August 2013 (UTC)

Done, thanks for the suggestion. Megahmad (talk) 19:33, 18 August 2013 (UTC)

Please see discussion at Wikipedia talk:Citation templates#New templates created, possibly redundant. --Redrose64 (talk) 12:54, 20 August 2013 (UTC)

I'd like to source, as an example, an information board at a castle near where I live in order to expand its article. Is this possible considering the text doesn't appear in a book or on the internet, and if so how would I go about presenting it? Thanks. Samwalton9 (talk) 17:24, 29 August 2013 (UTC)

The first question is: does it satisfy WP:V? Assuming that it does, you could use {{cite sign}}. --Redrose64 (talk) 16:03, 30 August 2013 (UTC)

I'd also factor in who wrote the sign, and whether it was a reliable source etc. If you don't mind me asking, which castle is it, btw? If its British or French, I might be able to help find something in the literature. Hchc2009 (talk) 16:07, 30 August 2013 (UTC)

I'll have to check who actually wrote the boards Redrose, not actually certain at the moment, haven't been there for a year or so! It's Dolwyddelan Castle, and though I have a few sources I'm planning to look up that my local library has in stock currently I'm sure the boards would contain a lot of good information. I'll head over and take a look at them soon. Samwalton9 (talk) 16:13, 30 August 2013 (UTC)

For Dolwyddelan, Richard Avent's work is the best - but it's only available in hardcopy that I'm aware of. There also are some good online sources available tough. Try the Royal Commission on the Ancient and Historical Monuments of Wales 1956 work, here (NB: the text and images are covered by Crown Copyright, which I believe recently expired on this volume). Prestwich's account on Google Books here would give you the end-history. You might also find this an interesting read. Hchc2009 (talk) 20:51, 30 August 2013 (UTC)

I just came to take a look at WP:CITEVAR because of discussion at another page. I noted that it mentions converting between paranthetical and reference templates. Is this just an outdated relic? I can't think of any article that uses paranthetical references, and, if one did, I would convert them on sight without even thinking about it. Hand writing a references section is obviously incorrect because it would mean that every new reference added would require a similar change to a references section (renumbering everything, or whatever). While WP:CITEVAR says that we shouldn't argue about, say, switching from citation templates to harvard references, I can't believe that anyone would actually support retaining paranthetical citations. Qwyrxian (talk) 23:02, 29 August 2013 (UTC)

Parenthetical references are used in some articles. They are also known as Harvard citations. They don't use numbers, so there would be no need to renumber citations when a new citation is inserted. Jc3s5h (talk) 01:23, 30 August 2013 (UTC)

Actuary. --Redrose64 (talk) 16:04, 30 August 2013 (UTC)

Be careful to distinguish parenthetical referencing, also known as author-date or Harvard style, from the {{Harv}} templates used for that and other styles. ~ J. Johnson (JJ) (talk) 20:12, 30 August 2013 (UTC)

I would definitely oppose you if you tried to change an article that has used parenthetical citations to citations in footnotes. Don't do it. That would be bad. Unless you can get consensus on the article's talk page. --Trovatore (talk) 20:26, 30 August 2013 (UTC)

There are a lot of people who believe that using 'little blue numbers' is required, so in my experience it's very easy to get a "consensus" to convert the style, even over the objections of the people who actually write the article.

In my opinion, PAREN is best suited to articles that cite a dozen different pages each from a small number of books (or long articles). Most of our articles cite a dozen different short sources exactly one time each. PAREN is not so well suited for that, especially for sources that have no author. WhatamIdoing (talk) 01:19, 31 August 2013 (UTC)

To me the most important benefit of parenthetical cites is that they allow explanatory notes to be very visually distinct from citations. It's true, you can distinguish them even with footnotes, [1] versus [nb 1], but it's not as clear a distinction as saying parentheses are citations, superscripts are notes. Explanatory notes are especially useful in technical articles, which are my cheif area of concern. --Trovatore (talk) 01:41, 31 August 2013 (UTC)

It's my impression that Harvard referencing is most suitable for scholarly articles where there are a limited number of authors involved and most readers will already be familiar with the author names and publications. --Robert.Allen (talk) 04:10, 31 August 2013 (UTC)

Yes, you may need to be familiar with author/date pairs if you use plain text parenthetical referencing. But if you use the {{harv}} template, as Actuary does, you have a blue link that takes you to the appropriate full citation. If you use Chrome, Firefox, Opera or Safari, an additional benefit is that the target citation gains a pale blue background so that you can pick it out easily (this doesn't work in IE because Microsoft haven't coded for the :target pseudo-class).

You could say the same things about Shortened footnotes. --Redrose64 (talk) 14:40, 31 August 2013 (UTC)

Even where there are a large number of authors "Harvard referencing" (i.e., author-date) is very handy for short citations (multiple references to a full citation), avoiding the limitations of named refs. ~ J. Johnson (JJ) (talk) 23:17, 31 August 2013 (UTC)

I'd never use it myself, but it is still around, and protected from drive-by "ooh, i don't like that" changes by CITEVAR like any other system. Some people seem to have difficulty grasping that the policy means what it says. Johnbod (talk) 21:10, 2 September 2013 (UTC)

Question. This guideline, and professional writing to my knowledge, both indicate that the proper placement of a reference is after the sentence (or, sometimes, at the end of the paragraph) of the material the ref supports.

I have been chatting with an editor who believes, for stylistic reasons, that it is better to -- where there are two paragraphs, and a footnote supports both -- only place a ref after the second para. Leaving the first para unreferenced. Even though, in this instance, it has been challenged under the policy wp:v.

He maintains that the ref after the second para is sufficient to confer referenced status on the first para.

(I've pointed out among other things that, this being wp, someone can easily at some point insert a para between the two, that is unreferenced, and we won't have any clue which paras that ref is meant to cover).

Any thoughts?--Epeefleche (talk) 23:54, 22 August 2013 (UTC)

My tendency is to cite after every paragraph (even if the citations are identical) for the reason you mention. Even so, the problem still occurs within a paragraph where more than one sentence describes the material cited, or even within a single sentence where there are multiple points to cite. I suspect there is no good answer, short of some kind of scoping template. (E.g.: "{{cited|The sun is big....|<ref> ...</ref>}}.") ~ J. Johnson (JJ) (talk) 21:39, 23 August 2013 (UTC)

The other editor is following a convention that is widely accepted in the academic world and noticeably unpopular here. See WP:MINREF and the sections that follow. There are no truly good solutions to this problem at the moment. WhatamIdoing (talk) 23:55, 26 August 2013 (UTC)

I should mention that the other editor feels so strongly about his position, that when I -- at a third editor's suggestion -- added footnotes to the unreferenced paragraph myself, the other editor deleted them.--Epeefleche (talk) 23:58, 26 August 2013 (UTC)

You should try to explain that in a crowd-sourced website, 2 paras is just too long to leave (often one is) - no one will be confident the reference covers that far back. NB the DYK/GA/FAR criteria, which all demand a minimum of 1 ref per para - I'm surprised to see MINREF says "Wikipedia does not have a "one inline citation per sentence" or "one citation per paragraph" rule, even for featured articles" which I don't actually think is true in those cases. Johnbod (talk) 21:13, 2 September 2013 (UTC)

GA requires nothing of the sort; GACN explicitly disclaims it. When I've asked at WT:FAC, the reply was that the goal was the fewest number of citations possible. DYK's supplementary rules page has a "rule of thumb" (not a requirement or a demand) for one ref per paragraph, but with a sizable list of exemptions. WhatamIdoing (talk) 05:12, 7 September 2013 (UTC)

Is there some sort of standard for how to cite source code in Wikipedia? Such as to cite a new feature when no changelog exists. Does source code even count as a valid source for citations? →Kyosuke Aoki∙♫_talk∙^contribs♪ 23:42, 6 September 2013 (UTC)

Source code could be a valid source, if it was published by a reliable source. However, it would be very easy for the editor adding it to interpret it, or use it out of context, in a way that is not supported by the original source, so it should be used with great caution.

Wikipedia does not have a standard about how to cite anything; follow the style already established in the article. Jc3s5h (talk) 00:38, 7 September 2013 (UTC)

The template {{reflist|refs= }} is part of the reflist documentation. It uses exactly the same format as {{reflist }} itself, but it places the citation and footnote info in the section where {{reflist }} appears. That means that when editing text, only the footnote designation <name/> appears, removing the distraction of having footnotes mixed in with the text being edited. An example of an article using this approach is Epistemological pluralism, and if the edit window is opened one can see how it works. The unfamiliarity of some editors with this approach leads to invocation of WP:CITEVAR to resist this change.

So here it the question: Can WP:CITEVAR be used to block the introduction of this simple variant of {{reflist }}, or is it meant to apply to radically different citation methods, like the Harvard approach? Brews ohare (talk) 16:45, 31 August 2013 (UTC)

• Possibly what this question amounts to is a discussion of whether {{reflist|refs= }} has advantages over {{reflist }}. I'd say that {{reflist|refs= }} removal of distractions in editing text is a significant advantage. It may be that the labor of setting up this approach is too great for some editors, but if another editor is willing to invest the labor, why object? I don't think that WP:CITEVAR is a policy designed to revert such a change. If a later editor wishes to use the other system, {{reflist|refs= }} is compatible with an inline <ref>Blah...blah...blah</ref>, so no editor is forced into anything. Brews ohare (talk) 17:30, 31 August 2013 (UTC)

• I believe CITEVAR does cover changing a system of citation in this way; even if you consider it is a better system, you should raise on the talk page page first. Hchc2009 (talk) 17:58, 31 August 2013 (UTC)
• I agree. This is a matter of personal preference: you may find it a distraction but other editors find it far easier to place references inline. It means for example when section editing, far easier in a long article, both the text and reference can be added at the same time. Easier to do and it means there's no intermediate unreferenced version. It also makes it easier to copy and paste content, when e.g. splitting a long article or otherwise rearranging between articles. And such a personal preference is not grounds for changing the article. WP:STABILITY also applies; although that does not mention citation style one of the arbitration decisions it is based on, Wikipedia:Requests for arbitration/Sortan#Preferred styles, does.--JohnBlackburne^words_deeds 19:02, 31 August 2013 (UTC)

Blackburne: The change of existing references to the format {{reflist|refs= }} intrudes in no way upon editors inserting new footnotes using an inline <ref>Blah...blah...blah</ref>. Their freedom is not impaired. So a revert back to {{reflist }} accomplishes only a negative result: it reintroduces extraneous citation information and (sometimes lengthy) commentary back into the text being edited. Brews ohare (talk) 19:11, 31 August 2013 (UTC)

Again it is simply your preference. Many other editors, including myself, prefer inline references. If you feel so strongly about this, and have good reasons (not just bald assertions like they are a "distraction" or "negative") then you could seek to change policy to recommend your preferred approach. But before doing so I would remember how your many other attempts to change policy have fared.--JohnBlackburne^words_deeds 20:00, 31 August 2013 (UTC)

Again, you miss the point. There is no interference with editors that wish to use the plain {{reflist }} instead of {{reflist|refs= }}. It is a zero disturbance to such folk. Brews ohare (talk) 23:35, 31 August 2013 (UTC)

No, you are missing the point. This is covered by WP:CITEVAR, as other editors here and at Talk:Pluralism_(philosophy) agree. So you should not be changing it simply to your preference as that's all it is. I prefer references that appear inline in the text, as I think do most editors given that they seem the predominant way of adding references to articles.--JohnBlackburne^words_deeds 00:01, 1 September 2013 (UTC)

• Actually it can disturb editors used to inline references that edit by section. LDR require editors to edit the entire page at once to maintain the relationship between sources and where they are used, which, for an article undergoing "rapid" changes, can be a hassle. It should be considered just like switching reference formats and thus covered under CITEVAR. --MASEM (t) 00:09, 1 September 2013 (UTC)

Masem: I have taken the liberty to reformat your comment with an asterisk because I think you have put your finger on what is the only reason for any editor to oppose conversion of references to a list-defined format: it can slow down 'hit-and-run' editing reversions for inexperienced rapid-fire reverters who don't understand their roll-back tools. Brews ohare (talk) 12:07, 1 September 2013 (UTC)

No, that's not what I said. It hurts the normal editing process for experienced editors used to in-line references, and can avoid edit conflicts since one only needs to edit by section, not the entire article. Typically, for myself, when I have a longer, more established article (50k of text, maybe 8+ sections), and some new references come up, it is very easily and very clear to edit the specific section the material that will be sourced with those references right there. With LDR one either has to edit the entire article at once (Which can easily lead to edit conflicts), or make two edits in a row, remembering how the references were named between edits. It is a drastically different editing workflow. LDR has their place and are certainly a fair alternative and one I deal with if I come into an article that uses them, but it is a function of editor preference and thus one that comes under CITEVAR. Thus, wholesale switches between LDR and normal inline cites (either way) should not be done without gaining consensus first. --MASEM (t) 13:12, 1 September 2013 (UTC)

(ec)Or to put it another way Brews, editors engaged in active changes to an article which they know will be contested can use it to make life difficult for editors who disagree with them. The obvious solution for any reasonable editor is to use inline until the changes are agreed and stable and then make the change. ----Snowded ^TALK 13:13, 1 September 2013 (UTC)

Masem: Thanks for the extended explanation of the advantages of avoiding list-defined references. Of course, the implementation of list-defined references does not interfere with adding in-line references, if that is your preference. One thing is very clear: unless list-defined references are introduced by the editor originating an article, application of CITEVAR means they will rarely be introduced because proponents of the in-line system will block any change, in my opinion, as much out of unfamiliarity as because of real problems. Brews ohare (talk) 13:34, 1 September 2013 (UTC)

Oh, don't get me wrong. If an editor wishes to start an article with LDR, CITEVAR applies to and that sticks. But yes, that always works in reverse - if editors are happy with inline cites, yes, you probably won't get them to switch. Both systems have benefits and problems, so it is silly to try to promote one over the other, and thus why CITEVAR is completely the right rule to apply. --MASEM (t) 14:33, 1 September 2013 (UTC)

• Almost unrelated comment "It may be that the labor of setting up this approach is too great for some editors, but if another editor is willing to invest the labor..." Much of the process of converting to LDR can be done with User:PleaseStand/References segregator. It can convert to and from LDR. More great tools can be found at WP:CITETOOL. Cheers. 64.40.54.89 (talk) 07:43, 1 September 2013 (UTC)

• "List-defined references" are little more than a clever way of using named refs so that the master "ref" (the "<ref name= ... />" entry), where the full citation with ALL the bibliographic details is often placed, can be extracted from the text. And that is the key issue: some editors prefer to keep their citation together with the text, others see that as obscuring the text and prefer to collect them in a separate area. LDRs are one way of doing the latter, but hardly the best. The radical aspect of just about any other method of doing this is the idea that full citations do not have to be in <ref> tags. Once you break out of that constraint you don't need to make reflist do grotesque things.

Masem: neither LDR nor any other system that segregates citations needs to be edited "entire page at once" to avoid broken ref links, etc. Just add the full citation to your References section first, and then the target is already there when links are added in the text. ~ J. Johnson (JJ) (talk) 18:06, 1 September 2013 (UTC)

Or use a browser which supports multiple tabs. In the case of Firefox and most other Windows browsers, you can right-click the [edit] link to open the edit window in a new tab. Do that in the refs section, and then you can edit the relevant section in the body in the same way. --Redrose64 (talk) 19:12, 1 September 2013 (UTC)

It still requires two separate edits - even if you do the tabbed implementation to prep the reference in one table. This can still be problematic on fast moving articles, and for at least my experience in doing major expansion of an otherwise stable article, slows down the expansion process. --MASEM (t) 15:52, 2 September 2013 (UTC)

So? How is a single complex edit better than multiple simple edits? Especially as whole-article editing is more likely to encounter an edit conflict, and then you have to extract all your changes and do them again. ~ J. Johnson (JJ) (talk) 20:25, 2 September 2013 (UTC)

If I'm just adding a text with a new reference to a single section, inline cites only require one edit with a low likelihood of edit conflicts (since the media wiki software is smart enough to handle simultaneous edits in different sections without conflict). --MASEM (t) 20:45, 2 September 2013 (UTC)

• The use of WP:CITEVAR. This discussion has changed my focus as to what the issue is here. It seems that using WP:CITEVAR to block implementation of list-defined references prevents its use, while allowing list-defined references has no effect upon those that don't like it, who can continue to use in-line footnoting unimpeded even after list-defined footnoting is implemented. Doesn't use of WP:CITEVAR this way stack the deck in favor of in-line footnotes for what is basically of no advantage to the in-line footnoter, a kind of dog-in-the-manger approach? Brews ohare (talk) 14:56, 2 September 2013 (UTC)
  • You are completely ignoring the valid complaints about LDR (eg the advantages of inlines over LDR). Again, both systems have benefits and both have problems, so there is no reason to try to force one over the other as the preferred method across WP as you are trying to do. --MASEM (t) 15:53, 2 September 2013 (UTC)

Masem: I think you misread me. I accept that there are preferences both ways; my point is that disallowing list-defined footnotes eliminates that particular option and its advantages, while doing the opposite and allowing list-defined references has no effect upon those that don't want to use this feature. So it's a dog-in-the-manger situation: I won't let you do what you want, even though it doesn't affect me." Brews ohare (talk) 16:10, 2 September 2013 (UTC)

Mixing different systems of marking up citations in the same article, for example, both list-defined references and references where the citation is contained between the <ref> tags, is inherently confusing and a bad idea. Jc3s5h (talk) 16:33, 2 September 2013 (UTC)

Brews, CITEVAR does not disallow LDRs, nor favor in-line cites; it says only that you should not switch (in or out) without obtaining consensus. There is no "stacking the deck" here, except that you think LDRs have no downside, and therefore should be favored. ~ J. Johnson (JJ) (talk) 20:28, 2 September 2013 (UTC)

The way you are talking, you are saying that you see the fact that CITEVAR prevents switching from inline to LDR that it is making it impossible to propagate LDR as a possible referencing mechanism, which is not true. The way that you are ignoring the benefits of inline references suggests you want to push LDR as the only right way to reference articles and that CITEVAR is a barrier to that, but the problem is assuming the project has agreed to move to CITEVAR, which is far from the case. Both systems presently co-exist, and like with the difference between reference formatting (eg harvard verse cite templates), shouldn't be moved without consensus in a specific article. --MASEM (t) 20:45, 2 September 2013 (UTC)

Perhaps this summary of the problem will help. (If it matters to you, I personally like list-defined references, but I recognize that this is not a universally held preference.)

• List-defined references are preferred by some editors because they reduce the amount of markup present in the editing window. This: <ref>{{cite journal | author=Westphal CF | title=Über einen merkwürdigen Fall von periodischer Lähmung aller vier Extremitäten mit gleichzeitigem Erlöschen der elektrischen Erregbarkeit während der Lähmung | year=1885 | journal=Berl. Klin. Wochenschr. | language=German | volume=22 | pages=489–91 and 509–11}}</ref> can be reduced down to the very brief <ref name=Westphal />.
• List-defined references are disliked by other editors because section-based changes require two edits instead of one. For example, if you remove a sentence and its citation, you must edit the content section and the reference section. Not all users realize that both must be removed, which results in ugly error messages on our pages.

If this change in formatting is controversial, or likely to become controversial, then it must be discussed on the article's talk page and the consensus respected. If, on the other hand, you're (truly) the only active editor at the article (e.g., you wrote all or nearly all of it), then I don't think you need to begin with a discussion. However, if someone objects to the conversion, then you should definitely not revert them until after you have established a consensus on the talk page. WhatamIdoing (talk) 21:00, 2 September 2013 (UTC)

Seems sensible. Anyway the answer to the original question " Can WP:CITEVAR be used to block the introduction ..." is clearly yes, as it will be for almost any change to a consistent and valid citation system. Johnbod (talk) 21:06, 2 September 2013 (UTC)

And CITEVAR can also be used to block the removal of LDRs.

WhatamIdoing's example can also be reduced to the nearly equally brief <ref>{{Harv|Westphal|1885}}</ref>. Which has the added benefit of not resulting in an ugly error message if it should be removed. ~ J. Johnson (JJ) (talk) 22:00, 2 September 2013 (UTC)

I believe that to have the Harv template work, you'd have to change the citation to use |last=Westphal and |first=CF rather than the current |author=Westphal CF. Beyond that, though, wouldn't that change actually produce shortened footnotes rather than a single full citation? Or do I have it confused with {{sfn}}? WhatamIdoing (talk) 05:21, 7 September 2013 (UTC)

{{Harv|Westphal CF|1885}} would work, but is incorrect in regards of the metadata; it should be first/last, Beyond that, I suspect you are thinking of {{sfn}}. The point I was making is that if the full citation is taken out of the text — which is the main point of doing LDRs — it can replaced with a Harv template (as I showed) as readily as a named ref (as you showed). ~ J. Johnson (JJ) (talk) 03:03, 8 September 2013 (UTC)

To be clear, J Johnson is absolutely correct if the article is standardized on LDR, then it is improper to switch back to inlines without prior discussion, same as with from inlines to LDR. --MASEM (t) 05:26, 7 September 2013 (UTC)

This discussion has taken the turn that WP:CITEVAR requires consensus for a switch to use of {{reflist|refs= }} if {{reflist}} already is in use, and vice versa. Of course, consensus is a wonderful thing, but the likely outcome of this interpretation is that the original choice will be preserved because consensus for a change will very rarely occur. So this interpretation is in practice a decision that no change is possible.

I call that the 'dog-in-the-manger' approach, because introduction of {{reflist|refs= }} doesn't stop anyone from using in-line footnotes if they want to, and permits the use of lost-defined references if that is wanted, on an editor-by-editor choice. On the other hand, use of {{reflist}} makes the use of in-line footnotes impossible.

So the more open and flexible approach is to use {{reflist|refs= }}, and I see no reason to use WP:CITEVAR as a means to insist upon the more restrictive choice of {{reflist}}. Brews ohare (talk) 16:35, 7 September 2013 (UTC)

While technically is is possible to mix inline and LDR systems, our MOS says to standardize on one or the other. The same could apply to any other CITEVAR issue (mixing Harvard vs cite templates, for example - there's nothing technically stopping us, but for avoiding editing wars, we standardize on one or the other). You're trying to push for a solution that is fixing something that is not broken in older articles. Going forward, I agree that LDR should be promoted, but for established articles, there's no need to force the LDR and if you can't get consensus to do so on talk pages, then just let it be. --MASEM (t) 16:40, 7 September 2013 (UTC)

An article generally evolves over time. In my experience, articles using list-defined references usually contain in-line references too, because some later addition was made by an editor who didn't want to bother with {{reflist|refs= }}. Possibly some later editor will convert this in-line version to a list-defined version and add the missing information that usually is missing in a hurried in-line footnote.

A mix-up is even more common with the Harvard referencing approach, because many editors have never learned how this method works either, and so they just put in in-line footnotes.

In summary, a mixture of footnoting is common, and it affords no difficulties for the in-line footnoter. And if a later editor wants to switch the in-line footnotes to list-defined footnotes, I have no problem with that, and see no reason to prevent changes by invoking WP:CITEVAR other than quarrelsomeness. Brews ohare (talk) 16:55, 7 September 2013 (UTC)

We're not talking about short term ref formatting inconsistency (it makes no sense to slap a new editor's hand if they add a bare url citation to an article that is otherwise consistently in one style, for example), we're talking at the end of day quality/formatting one would expect at an FA/GA assessment. To that extent, adding an inline citation to an article that is already LDR can easily be fixed into an LDR, but the reverse, introducing LDR to an article with inline takes excess work and can be seen as submarine-introduction of LDR. End of the day, there is no need to push LDR over inline (or inline over LDR), and we let CITEVAR deal with avoiding edit wars. --MASEM (t) 17:02, 7 September 2013 (UTC)

Well, let's look at it this way. Editor A wants to put in the effort to change all the in-line references to list-defined references; Editor A just finds all the digressions in the edit window makes it difficult to follow the exposition. Now, Editor B says they object to change. Editor A asks "Why do you object". Editor B says "Because WP:CITEVAR says I can object, and don't need any reasons."

Does this exchange make any sense? Is invoking WP:CITEVAR aiding the formation of consensus? Is policy used like this improving WP? Brews ohare (talk) 17:10, 7 September 2013 (UTC)

If editor B is a primary or original editor of the page, yes. In any other case, if B objects, then A should propose on the talk page to make the switch. --MASEM (t) 17:18, 7 September 2013 (UTC)

Three questions were asked. While 'yes' possibly can be applied to Q1 depending upon your taste, the answer to Q2 and Q3 is clearly 'no': Use of WP:CITEVAR does not promote discussion of pros and cons, and consequently does not allow a sensible choice between alternatives. In this regard, application of a policy without any attempt at explanation is just frustrating discussion, and unfortunately is a common use of WP policy. In this particular case, in fact the question could be settled once and for all because use of {{reflist|refs= }} denies no-one use of in-line references if that is what they want to do, and so using WP:CITEVAR to prevent use of {{reflist|refs= }} is counterproductive in restricting editing flexibility. Brews ohare (talk) 18:42, 7 September 2013 (UTC)

WP works by the mantra, "if its not broke, don't fix it". There is no need for established articles to switch from inline to LDR. In new and recent articles, yes there may be benefits for starting with LDR or suggesting it and that's where if one wants to make the effort to get more editors using LDR, to use it there. But just as it makes no sense to try to get editors to switch from harvard to cite templates or vice versa on an article long-established on one, it doesn't make sense to do the same with inline vs LDR, unless you can argue there is a real benefit to switching over specific to that article (eg, we all know that LDR is more newer-editor friendly, but that's not a specific benefit to the article in question). And again, we should not be knowingly mixing inline and LDR - it's different for a new editor to come in, unaware of the situation, and use the wrong citation approach, and for an established editor to ignore the established approach and insert the wrong style in an article. --MASEM (t) 18:51, 7 September 2013 (UTC)

Do we actually strongly discourage mixed use? I could imagine editors deciding to use LDR for any ref that gets re-used, but leaving the rest inline.

As for "the effort to change to list-defined references", the "effort" is something like clicking three buttons if you've got the script installed (which I recommend), so it's not really much effort. WhatamIdoing (talk) 05:13, 9 September 2013 (UTC)

This edit request has been answered. Set the |answered= parameter to no to reactivate your request.

From 3.1.1 How to create the list of citations: "This section, if needed, is usually titled "Notes" or "References", and is placed at or near the bottom of the article ." Can we delete the extra space before the dot? 81.232.114.228 (talk) 20:30, 11 September 2013 (UTC)

Yes, we can! And done. ~ J. Johnson (JJ) (talk) 22:04, 11 September 2013 (UTC)

Resolved

I frequently use information from the Paleobiology Database (PBDB) regarding the locations where a particular fossil was found, see this example for Lonchodomas carinatus. I think it has value to cite the original source of the information, but I do not think I need to read the article itself because the information is in PBDB. It would be very helpful if the citation template would have an option to include |cited in= |, that would generate that text the reference list. I'm only interested in the solution of my problem, not in the programming, so I hope my issue is considered valid, worth solving and that someone will create what I think I need. Dwergenpaartje (talk) 09:42, 15 September 2013 (UTC)

Well, you are essentially providing two citations. The simplest solution is simply to use two citation templates and put them between ref tags i.e. <ref>{{cite book [details1]}} Cited in {{cite book [details2]}}</ref>. Betty Logan (talk) 11:25, 15 September 2013 (UTC)

Thanks, I'll give it a try! Dwergenpaartje (talk) 19:05, 15 September 2013 (UTC)

There is plan to use a bot to trim cruft from Google Books URLs in citations. We need someone familiar with their format, to advise on which parameters can safely be trimmed, and which, if any, should be left. Can anyone advise, please? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:37, 17 September 2013 (UTC)

I link them a lot, and the only parameters I have found that need to be retained are the "pg=", the "v=" and "q=" parameters. Stripping out everything else after the book id number seems to have no effect. I have added hundreds of links to pages/word searches on Google Books and these are the only parameters I have ever used to my recollection. If you want a sample set with a high concentration of such links then I have added dozens at List of highest-grossing films. Betty Logan (talk) 13:02, 17 September 2013 (UTC)

Where there is no specific "pg=" then the link can access by text parameter "q=" for matching text on various pages, as two very different forms of link. -Wikid77 (talk) 12:33, 20 September 2013 (UTC)

I think that there are four basic types of google book cites:

1. a cite of a particular page where the book may be previewed
2. a cite of a particular passage where the book is limited to snippet view
3. a cite that links to the book's cover (preview mode)
4. a cite that links to the book's summary page

When I've encountered type 1 with a lot of cruft, reducing the link to use only the page parameter works well.

For type 2 cites, very often the snippet view will show some of the search terms but not enough to support the statement in the article text. I tend to remove snippet-view links unless they clearly support the article's text. That happens rarely enough that I can't remember ever trimming the url.

For type 3, I simply remove everything but the page parameter.

For type 4, if the book has preview mode I'll add the appropriate page parameter; if no preview mode, I remove the url from the citation.

I test all of these new cite / link configurations before actually making the change.

—Trappist the monk (talk) 13:13, 17 September 2013 (UTC)

Er, why are we linking to google books for this? The problem with GBooks' previews is that the function is inconsistent around the world and even sometimes within the same region. A cite to the book with page number is more than sufficient and while this can provide the URL field for books, seems far too haphazard to consider reliable. And if one needs book information, just including the ISBN is sufficient to get to the list of pages that provide details on the book. (Note that I'm not saying one can't use GBooks to find information and then cite it appropriate with a cite book template, but you don't have to "show your work" to include the URL To the pages). --MASEM (t) 13:24, 17 September 2013 (UTC)

I view google books links as merely a convenience for the reader. This is, in my mind, no different from citing The New York Times or any other news source and providing a link to the online copy of the article. Articles can disappear behind paywalls, be deleted, etc. just as google can make a book's page available or not. If it's available, why not provide a convenience link?

—Trappist the monk (talk) 13:39, 17 September 2013 (UTC)

Because it doesn't work for 100% of the readership - contents of books may be blocked by region, or depending on which way the google algorithm spins, certain pages may not be available to some. It would be different if the full contents of the book were available for all through google, at which point I'd be perfectly okay with that, but we're talking that while a useful resource is not one all of our readership can use or access. --MASEM (t) 14:18, 17 September 2013 (UTC)

I am with Masem here, why link to Google books. We have ISBN for a reason, and that does not favour anything (including the search ranking of books by telling google which books are regularly read by Wikipedia users). I know, it is a convenience to have a link directly to the page, but it does favour one 'host' over others while there are methods to avoid that. One could even consider to disable the url parameter when there is an ISBN or other general search available in the template (at that point, the url can always be copied/pasted from the editor when an editor needs it, readers can make their own choice through the ISBN) ...

(after ec) Trappist the monk, the problem is that it favours one source over the other. For the New York Times it is published by them, for Google Books .. hardly any books are published by them, they are not even the copyright owner on most of them (they show pieces under fair use or with permission ..). Similar services are given by many bookshops, amazon, etc. etc. And I think that that is an important difference. --Dirk Beetstra ^{T C} 13:44, 17 September 2013 (UTC)

The problem though is that if you have sourced the information from a Google Books preview then the citation should ideally state that. It may well be identical to a hardback version with an ISBN and all that, but the bottom line is if you obtained the claim from the preview version you should cite the preview version. Betty Logan (talk) 14:16, 17 September 2013 (UTC)

There's no requirement for that, working on the assumption the Google book scan is an exact copy of the reported printed version (include print edition #). Unless you can prove that Google is altering books they have for preview on a regular basis, the source remains the book itself, not Google Books, and anyone can verify the claim by getting the book. (Of course, if it comes down to editors disputing a claim, you can toss the Google Books link up on the talk page to show that, but that process should be invisible to the reader). --MASEM (t) 14:21, 17 September 2013 (UTC)

Google Books Preview is essentially no different to an online newspaper article, which may be behind a paywall or inaccessible in certain countries. WP:SOURCEACCESS doesn't say that online sources have to be universally accessible, but WP:SAYWHEREYOUGOTIT does state we should say exactly where we got the content from. I see this as no different to online newspaper articles really: there may well be a hardcopy version, but if you got your information from the online version that is the version you cite. Betty Logan (talk) 14:25, 17 September 2013 (UTC)

It is very different, in that Google Books in of itself has no affliation with the book, compared with newspapers or academic papers. Further, I point again to the fact that what is available in a Google Books previous is inconsistent across the readership. This is a difference from PAYWALL/SOURCEACCESS is that anyone that had the proper credentials or purchases access will always get the same, full content from the source. We simply can't assure that for Google Book Previews, and thus doesn't make sense to use them for citations. --MASEM (t) 14:36, 17 September 2013 (UTC)

Betty Logan (talk · contribs). Say where you got it. You got it from the book. You find it nevessary to say that the specific copy you used was on Google books ... I have no-one seen writing 'from my bookshelf' as where the specific copy of the book was.. You got it from the book, attributing it to google as a source is untrue, google did not publish it. An Google preview is not the same as doi, jstor, isbn, as those are really pointing through to the original publisher. Those are proper links, google/amazon and wayback are pure convenience links which often have nothing to do with the original publisher. And that is the importance of saywhereyougotit, to acknowledge the original publisher. --Dirk Beetstra ^{T C} 04:50, 18 September 2013 (UTC)

It seems to me there is no reason to spurn a free look at a page in a book just because it isn't available to all Wikipedia readers, any more than one should spurn a paper book because some local libraries have it on the shelf and others don't. Jc3s5h (talk) 14:35, 17 September 2013 (UTC)

Agree with User:Jc3s5h. If the GBooks URL works for the person adding the reference I see no reason why it can't be included. Compare the 'jstor=' and 'mr=' options in the {{cite}} template, which provide ways of reaching article text or abstracts that are only available by subscription but are still useful. EdJohnston (talk) 23:29, 17 September 2013 (UTC)

No, it's different. JSTOR/MR links are assuredly going to give me the full source regardless when and where I access it - as long as I have access credentials there, or in most cases, can purchase those. Google book previews are random and inconsistent even if free. Further, why should we favor Google over any other service (like how we handle ISBNs using a landing page that points to dozens or ways to get book details) such as Amazon's preview feature? It may be a convenience at times, but at other times will be misleading the reader. --MASEM (t) 23:49, 17 September 2013 (UTC)

Agree with Jc3s5h. I find Masem's objections unpersuasive. ~ J. Johnson (JJ) (talk) 00:47, 18 September 2013 (UTC)

jc3s5h, noone is spurning the citation, isbn will still lead you through to google books, as well as other online sources who may offer previews. --Dirk Beetstra ^{T C} 04:50, 18 September 2013 (UTC)

I would tend to agree with jc3s5h... As a reader I find it hard to a imagine a situation where I would encounter such a link and wish it weren't there. Seems the worst that happens is, it doesn't work/I can't access the content - which will be true of many other links, e.g. most all subscription based sites. Christopher Parham (talk) 05:34, 20 September 2013 (UTC)

We have a problem with readers who go to a link, find that it doesn't work for them, and thus remove the entire citation to a book (rather than the just a dead URL). So it is a risk, albeit IMO a small one. WhatamIdoing (talk) 15:20, 23 September 2013 (UTC)

I think that would be very unusual - could happen though. I would expect someone to be watching the page, spot removal of a cite and reverse it. I would prefer to leave the url if it does go dead, just in case some way is found to recover it later. A different argument in favor of urls, which a couple of students have mentioned to me when discussing the accuracy of Wikipedia, goes something like: "I find Wikipedia is a good starting point, anyway, and the list of sources at the back is useful. I can click on those links and find out more about the subject." Aymatth2 (talk) 16:21, 23 September 2013 (UTC)

My take on this: (1) Editors can adequately verify a source by reading enough of it however they can. (2) Providing a link to an online page view, Google or otherwise, is a courtesy that will hopefully be useful to other readers. The important thing is that the book is the actual source, not the online view of it. The argument that the online view is inconsistent in its availability is not convincing, since the online view is not the source but just a potential way of looking at the source. Note that Google snippets are very dangerous as vital content can be invisible; often they will fail step (1). Zero^talk 01:05, 18 September 2013 (UTC)

• Disagree with Masem. Whilst not everyone can see them, google previews seem mostly stable of a period of years, and often allow a whole chapter or paper to be read by anyone in the right area. Including them as a supplement to the usual bibliographic details is highly useful. I don't see any point to linking to non-preview views, & snippets are rarely worth linking to (or using as refs). Johnbod (talk) 11:57, 18 September 2013 (UTC)
• This discussion seems to have strayed from the original topic. Views:

Why link? When citing a book, article or paper I describe the source and say where and when I found it, naming authors, title etc., url and accessdate. This applies to JSTOR, Google Books or any other website. The url helps another editor or reader to check what I have added. They might find a paper version in the library, but if the source is available online and they have access to it, that is easier. Aymatth2 (talk) 12:55, 18 September 2013 (UTC)

How to link. Like most websites, Google Books is far from perfect. I have found dead links in the search results list and pages that were visible but now are not. Occasionally the description of one book is combined with scanned pages from another. The database changes. For this reason, I provide both url and accessdate, as with any website link. Maybe, some day, Google will provide a service to access their archived database as of that date. I use the Wikipedia citation tool for Google Books to build book descriptions, place them at the foot of the article and link to them via the {{sfn|author|1913|p=123}} template. The idea is to provide all information that may possibly be relevant to someone checking, but to place it where it does not distract casual readers. Aymatth2 (talk) 12:55, 18 September 2013 (UTC)

Trimming cruft. For a book page visible online, anything after the page number can be dropped. E.g. http://books.google.com/books?id=E7EyAQAAMAAJ&pg=PA1002. If the full page is not visible (in rare cases, snippet view does clearly support the information), or if no particular page is being referenced, as in a bibliography of books written by the subject of the article, I would just give the url up to the first ampersand, as http://books.google.com/books?id=h-PdQAAACAAJ. The other parameters are just too inconsistent and unreliable. I can see no risk in a bot that trims all parameters other than &pg=. Aymatth2 (talk) 12:55, 18 September 2013 (UTC)

If you drop the "q=" you can lose snippet information. For instance, http://books.google.com/books?ei=CcB9UPf2IOSj0QXR-YH4DQ&id=YSBXAAAAMAAJ&q=%24142+million is used to support the fact that The Godfather earned $142 million. It is not a particularly contentious fact, and while the snippet does not allow for much context it is adequate to convey the fact. However, you need the "q" parameter to access this factoid, so dropping it would have repercussions for many sources that use this facility. The only parameters that should be dropped should be those that have no bearing on the accessing of the information. Betty Logan (talk) 13:03, 18 September 2013 (UTC)

I don't feel strongly on this - the example is a good one. If I give just the url as http://books.google.ca/books?id=YSBXAAAAMAAJ (but cite the page number), an editor checking can enter "$142 million" in "From inside the book - Search" and find the factoid. Think I am talking myself into saying that we should also keep the q parameter. The minimalist format for the bot would be http://books.google.ca/books?id=YSBXAAAAMAAJ&q=%24142+million. "ei=" is not needed. With an accessdate, of course, in case Google starts to give different results later. Aymatth2 (talk) 13:21, 18 September 2013 (UTC)

Remember that you are citing a book with a publication date. Because of the not-always-available-to-all characteristic noted above, a google books link is merely a courtesy for readers and not the cited work. As such, |accessdate= isn't required and contributes little if anything to the citation.

Often, the &q= parameter contains search terms that don't reflect the tidbit of information being cited. This I think can be confusing to readers who follow the link. For snippet views, it is important to keep the &q= parameter if the snippet view clearly shows the cited information in the appropriate context (and isn't a simple case of text matching the search terms). When preview is available, the highlighted search terms can be and, in my experience, often are distracting.

Can a bot tell the difference between a snippet view and page view simply from the url? If it can't, then I think that there is nothing much for the bot to do. I don't know the answer to this question.

—Trappist the monk (talk) 13:56, 18 September 2013 (UTC)

The book definition gives two different types of information: 1) the work that contains the information (author, title, edition, page, ISBN etc.) and 2) the website where a copy of the work was found, with the date when it was found. The website could be Google, JSTOR, Project Gutenberg, whatever. A url is a fragile identifier because the website may be shut down or the website owner may change what the url points to, so it should always be accompanied by the accessdate. Some sort of archive service may be able to recreate what that url pointed to on that date. I believe this will become increasingly relevant as we move away from paper versions of books, papers etc. Aymatth2 (talk) 14:44, 18 September 2013 (UTC)

Editors can and probably will mess around with the url. In the example above, they might change http://books.google.ca/books?id=YSBXAAAAMAAJ&q=%24142+million into http://books.google.ca/books?id=YSBXAAAAMAAJ&pg=PA32 in an attempt to go to the page they see in snippet view. It won't work, but the bot can't tell that. Still, if the bot drops all parameters except id=, pg= and q= it is probably safe enough. Just needs some experimentation to confirm. The search terms could well be different from the factoid extracted from the snippet. I don't see that as a real problem. Only certain types of people would bother to check links. It seems better to let them see what showed up, even if the search term is a bit distracting, than to not show them anything. Aymatth2 (talk) 14:44, 18 September 2013 (UTC)

• What about retaining the parameter "v=" along with the others? -Wikid77 (talk) 12:33, 20 September 2013 (UTC)

• Good question. Yes, v=onepage does seem useful. In a sense, there is not much point having a bot change the url. It does point to what the original editor saw, and to the regular reader it is hidden behind the book or chapter title. Three reasons I can see for stripping down the url.

1. If the url in a book definition is inline to the text, as opposed to in a source definition at the end of the article, editing is a bit more awkward. A shorter url would be easier to work with
2. Parameters like &hl= may affect the display in unwanted ways, overriding the viewer's default
3. There may be some compromise of privacy, some information an editor does not realize they are giving out.

We need someone who really understands all these parameters. Not me... Aymatth2 (talk) 14:45, 20 September 2013 (UTC)

It is entirely conceivable that an editor may stipulate a two page view, but if "v=" is dropped it just defaults to the one page view, so the content is still accessible. Now, if the editor has specified the page numbers in the citation that clearly isn't a problem, but if they haven't and have just treated it like a web link you could fundamentally alter the reference so that the exact location of the information is no longer stipulated. Personally I would retain it for this reason; you cannot assume that every editor who uses Google Books has provided complete citation details. We have to account for human error, sloppy editing etc. If you want to err on the side of caution, the "pg", "q" and "v" links all need to be retained, or at least not removed by an automatic process. Betty Logan (talk) 15:01, 20 September 2013 (UTC)

An editor may provide a stripped-down bare url like http://books.google.ca/books?id=Cq22AAAAIAAJ&q=%22Louise+Saumoneau%22 giving snippets from three pages in the book. Not good practice, but the information is there for someone who wants to check the article. I am o.k. with saying we keep "id", "pg", "q" and "v", and drop all other parameters. Aymatth2 (talk) 16:47, 20 September 2013 (UTC)