Wikipedia talk:Citing sources

Archives

Source code citations?

Can source code or a comment inside source code be used in a citation? To me, It seems like it should definately be okay if its freely open source. But what if I want to cite something... not completely closed but that comes with a piece of software or book that requires purchasing? For instance, library source that are included with Visual Studio. Can this be acceptable or is it not valid because it cannot be checked be "everyone"? AllUltima (talk) 05:39, 21 December 2007 (UTC)

Things that must be purchased may be cited; the most common example is books. However, software is often protected not just by copyright, but also by a license agreement, so take care to obey your license agreement. --Gerry Ashton (talk) 05:45, 21 December 2007 (UTC)

Just to amplify what Gerry Ashton said -- Sources must be "available" and being available for purchase is fine. Inserts that come with books are therefore fine. As to license agreements, personally, I wouldn't worry too much about licensing terms for books that are sold on the consumer market: citing to content, factually described, within a license agreement would rarely violate the terms of most such licenses. That said, do be careful about stitching together cited information to form a new argument -- wikipedia doesn't publish original research. --Lquilter (talk) 14:06, 21 December 2007 (UTC)

Blacklisted links

This dead link check which I did today turned up several links which it flagged as "Blacklisted link". What is the recommended action here? Could info on the recommended action be added to the project page? -- Boracay Bill (talk) 07:00, 21 December 2007 (UTC)

Excessive link length that makes the page scroll horizontally

http://en.wikipedia.org/wiki/Friendly_fire#Notes

See reference 2, ^ http://www.publications.parliament.uk/cgi-bin/newhtml_hl?DB=semukparl(cont.) —Preceding unsigned comment added by 121.7.100.240 (talk) 16:29, 22 December 2007 (UTC)

External links should usually be named with the method at Help:Link#External links. A very long link like your example should definitely be named. The second best option is to just put it in brackets with no name, producing a numbered link like this: [1]. PrimeHunter (talk) 17:42, 22 December 2007 (UTC)

I hasten to draw attention to the following from the "External links" subsection of this project page: "As with all inline citation methods, a full reference would also be required in a "References" section at the end of the article". -- Boracay Bill (talk) 18:53, 22 December 2007 (UTC)

At least part of that link instructs the browser to highlight the words "friendly" and "fire" in red. If you can find those parameters and eliminate them, you can shorten the link without significantly reducing its effectiveness. --GentlemanGhost (talk) 18:42, 23 December 2007 (UTC)

Citing sources/example style sub-page

Regarding the Wikipedia:Citing sources/example style sub-page, another editor asked on its talk page about its status, but has been unanswered since January. Is the sub-page considered a "style guideline" as well? Or is it an "essay"? Or neither? --GentlemanGhost (talk) 18:35, 23 December 2007 (UTC)

Clarification on author-page

Harvard referencing here refers to the author-date system (Jimbo 2007), and does not mention the author-page system (Jimbo 23). Does this mean the author-page system is forbidden on Wikipedia? Gimmetrow 23:45, 26 December 2007 (UTC)

The article Harvard referencing makes it clear that the page number(s) are included in situations where they are useful. If you wish, you can think of it as the "at least the author-date system". --Gerry Ashton (talk) 01:07, 27 December 2007 (UTC)

If you are talking about footnotes of the form, "Author, page", where the authors refer to the list of references where complete information is found, that format is widely used and acceptable. I haven't seen people doing this in parenthetical references, but I don't think there would be great objections to it. Christopher Parham (talk) 01:11, 27 December 2007 (UTC)

Funny you should say that, but I had objections to it, so that's why I'm asking. Author-page means parenthetical citations of the form (Jimbo 23), where the 23 means p.23 of the work by Jimbo fully specified in the references. Nothing else appears in the parenthetical citation unless needed to distinguish two works by the same author. The year of publication appears in the references. This approach has some benefits over author-year for online work: websites, having no page, are identified by author only. For books, it's basically the "short note" system used in many FAs.

So the question is, does Wikipedia allow only that narrow form of parenthetical citation as described here, or more generally, any parenthetical citation system? Gimmetrow 02:36, 27 December 2007 (UTC)

Looking at the page you seem to be talking about, it seems to be fine except that I do not understand the reason the references are numbered and ordered the way they are -- alpha by author would seem to make more sense if the citations are pointing to an author name. In my experience the how-to examples given on this page should be viewed as indicative examples of acceptable systems rather than an exhaustive collection of the allowed options, and the style used on that page seems to be within the acceptable universe. There are other misconceptions floating about on that page, including the notion that the use of citation templates is a standard practice. Christopher Parham (talk) 02:51, 27 December 2007 (UTC)

New uw template?

I've seen a few times people deleting web sources which go dead and often deleting the information soon after as unsourced! I've just read [2] which says you shouldnt do this.

Could someone put together a user warning template I can use to respond to a user who deletes a dead link without replacing it? Thanks AndrewRT(Talk) 18:31, 27 December 2007 (UTC)

I'd say you're the man for the job, if you feel this is important. I would probably just drop that link on their talk page and ask them to keep it in mind. Christopher Parham (talk) 14:52, 28 December 2007 (UTC)

Done (by some other kind editor!) - see discussion at [3] AndrewRT(Talk) 18:05, 28 December 2007 (UTC)

A bit confused with the citation system, what to do and how to do it?

I'm trying to add references to an article I'm working on and I have difficulties to understand what to do and how to do it at best. After a few investigation about how Wikipedia works with respect to referencing, I'd like to share a few remarks . I apologize in advance if this is redundant with some other parts of the wiki.

• About the Wikipedia:Citing_sources documentation:
  • I did not find the documentation about {{wikiref}} and {{wikicite}} very clear. It took me a lot of time to understand that the first one is used for refering from the article body to a citation of the reference section and the other one to actually give the text of the reference in the reference section. At the beginning, I thought that one was an obsolete version of the other. I believe that it would help to reformulate the sentence "Links can be created using {{wikiref}} and {{wikicite}} ..." so that it is clear that one is expected in the body of the article and the other in a reference section of the article. The example shown is misleading as it does not show the usage of {{wikiref}}. It should certainly be said somewhere that {{wikiref|id=foo|text=this}} and [[#Reference-foo|this]] are (if I understand correctly) equivalent.
  • Still about {{wikicite}}, an example showing that it can effectively be used in combination with {{citation}} would be helpful I think.
• As a LaTeX user, I'm used with the BibTeX sytem and I wonder how to do similar things in Wikipedia:
  • The BibTeX system has two properties that I could not find in MediaWiki:
    1. It automatically orders the references in alphabetic order: no need to think about where to put the next reference in the reference list.
    2. It allows sharing of bibliographical notices: the same book, the same article of a journal, the same URL can have its {{citation}} template once for all in a separate repository and all what has to be done in the body of the article is to use a <ref name="the_canonical_identifier_of_the_reference"/>.
  • Of course, this would need some software support from MediaWiki, but we could then imagine the same architecture as it is done for inserting image. If a reference does not already exist in the common pool of references, we shall add it manually to the database (in the same way we already do for images). Then, to make a reference, the editor would simply have to give the identifier that comes with the reference. The selection of the necessary references and the sorting in the page would then take the place at the {{reflist}} template.
  • With such a kind of system, we could get rid of the two-step referencing that links first to a footnote before linking to the actual bibliographical data.
• Even without going to a system à la BibTeX/LaTeX, can we imagine having a global Wikipedia repository of references (Note: I also asked the question at the village pump and had interesting answers).
  • For instance, some robot could collect and classify all {{citation}} used in pages so that one could easily find the citation that we are looking for and that we know for sure that it is has been used many times already (but where?). This would help also to improve the overall consistency and accuracy of the bibliographic data used in Wikipedia (at least those used several times).
• Is the {{citation}} template enough general? Here are typical examples:
  • For an article I'm working on, I want to add a reference to a 1958 book by Curry and Feys with two sections by William Craig. This reference is already on the page about Robert Feys but it is plain text and it does not use the {{citation}} template. How can I say with a template that William Craig is an auxiliary author who only wrote two sections?
  • I have a similar problem with the following reference: Girard, Jean-Yves (1987-90). Proof and Types. Translated by and with appendices by Lafont, Yves and Taylor, Paul. Cambridge University Press. How can I explain, using the {{citation}} template that two of the authors only contributed to the translation and to appendices? In BibTeX, there is an optional field note for this kind of thing. It is not the perfect solution as it lacks robustness (it may depend on the actual format used for rendering the citation), but it helps a lot.
• Finally, why isn't there a similar mechanism of referencing as the Harvard referencing system provides but using a footnote instead of the Harvard (Author, date) format. This would provide another way to avoid the two-step referencing mentioned above. With such a system, each wikicite would provide a number (its position in the list) and each {{wikiref}} would refer to the corresponding number. Hugo Herbelin (talk) 17:59, 29 December 2007 (UTC)

(begin insert) Regarding Brenton, FWICS Frey, Curry, and someone else whose name I cannot read were co-editors. I haven't seen the actual book, but would guess that individual chapters were written by different authors. Perhaps cite a chapter (using {{Citation}}) something like:

• Brenton, Frederic (1969), "Some chapter written by an individual author", in Curry, Haskell Brooks; Feys, Robert (eds.), Combinatory Logic, North-Holland Pub. Co. {{citation}}: Check |chapter-url= value (help); Cite has empty unknown parameter: |1= (help)

Regarding Girard I think you need to append such explanatory notes after closing the body of the {{Citation}} template.

Finally, try looking at the documentation for {{Ref label}} and its companion {{Note label}}, and at Wikipedia:Footnote3. -- Boracay Bill (talk) 01:41, 30 December 2007 (UTC) (end insert)

{{Citation}} is a fairly general citation template. {{Cite book}} is a more specific template. It supports a "coauthors" field and an "others" field. The "others" can work like a note about translators. But, you don't need to use these templates anyway. Just write the reference like you want. As for LaTeX, many editors have asked for a BibTex-like structure, but it doesn't really exist. About the best you can do is use named references as you illustrate above. You just need once instance in the text somewhere that defines what the named reference means. See WP:Footnotes for more detail on that. Gimmetrow 18:29, 29 December 2007 (UTC)

I don't say that BibTeX has only advantages, but being able to have the bibliographic data at only one place is great. By this way, when you add an information you did not have previously (like the ISBN, the DOI, or the exact range of pages for a conference's article), it is changed at every place in the same time.

In the article I was talking about (Curry-Howard), I finally could manage to reuse the uniformity of style that the {{citation}} template provides without losing information: I simply added the missing informations after the citation. That worked because {{citation}} adds no period (note however that the field "other" was not displayed by {{citation}} and I had to add it manually).

At the village pump, I learned that French wikipedia has the "numeric Harvard" style I was talking about. I liked it very much (the example given was [4] but it seems in fact that all pages have the automatic indexation of <ref>'s directly within <references/>).

At the village pump, I also learned about the Zeteo database (for mathematics). It is also very convenient. Shall I consider that using this kind of tools (Zeteo, but also the Reference Wikification tool) is a recommended practice for dealing with citations?

---

• About the Wikipedia:Citing_sources documentation:

I've expanded the documentation to hopefully better explain how [[#Reference-...]] is effectively an easier to use 'free-format' version of {{wikiref}} template, which hooks up with {{wikicite}}. --SallyScot (talk) 19:22, 29 December 2007 (UTC)

Thanks. Then, if I follow the example, the recommended form is to use [[#Reference-...]] rather than {{wikiref}}, right?. But I'm still a bit confused because I thought that the advantage of {{wikiref}} over [[#Reference-...]] was that you did not have to know that the internal encoding of a link is the id of the citation prefixed with Reference-. Hugo Herbelin (talk) 21:49, 29 December 2007 (UTC)

The {{wikiref}} template page suggests that the [[#Reference-...]] format is simpler, but I guess it would be up to individual editors if they thought otherwise. I'd say [[#Reference-...]] is simpler because you don't include the parameter names (id=,text=) that way, but really I prefer it for it's flexibility. Using the full {{wikiref}} template obliges the citation to display with brackets, which is fine for Harvard referencing, but you wouldn't want the brackets for short footnote citations with full references. So [[#Reference-...]] is used in the both the examples (i.e. with Harvard referencing and with short footnote citations with full references), showing that the short citations method is essentially the same as Harvard referencing, basically replacing parenthesis with <ref> tags. --SallyScot (talk) 00:59, 30 December 2007 (UTC)

Continuing to understand better, but still finding that the Citing sources guideline page can be improved.

Let me try to summarize. I have a new article, I have to choose a reference discipline for it, and I have 4 parameters to take into account:

1. Are the full references scattered in the source of the page or are they gathered in a well delimited part of the source?

2. Do the references appear as footnotes of the form ^[1] or as text of the form (Miller 2005)?

3. Are the references directly linked to the full reference list or indirectly via a note list, or not linked at all?

4. Do the full reference list support references that are not linked from the body of the article?

Then, I have the choice between 4 citation disciplines which each provides the following features:

	Position of references in source	Reference style in article	Linking to full reference	Ability to have non linked references
Harvard referencing	all together	(Miller 2005)	direct	yes
short footnote citations with full references	all together	[1]	via note list	yes
footnote referencing	scattered	[1]	direct	no
embedded links (not recommended)	scattered	[1]	none	yes

Is that right?

Then,

1. if I want to gather the references alltogether (column 1), {{wikicite}} is what I need.

2. if I want a footnote (column 2), <ref> is what I need and <references/> makes the automatic ordering of the footnotes

3. if I want to directly link to the full reference (column 3), [[#Reference-id| ... ]] is what I need.

Is that right?

Especially, if I want Harvard citation, I use ([[#Reference-id| ... ]]), with parentheses around, for which there is an alias named {{wikiref}}.

Now, can we combine direct links and footnote? If one tries:

The Sun is pretty big [[#Reference-idMiller2005|^[1]]], however the Moon is not so big [[#Reference-idSmith2006|^[2]]].

1.{{wikicite|id=idMiller2005|reference=Miller, E (2005). "The Sun", Academic Press.}}</nowiki>

2.{{wikicite|id=idSmith2006|reference=Smith, R (2006). "Size of the Moon", Scientific American, 51(78).}}</nowiki>

Then, it actually produces the correctly linked following text:

The Sun is pretty big^[1], however the Moon is not so big^[2].

1. Miller, E (2005). "The Sun", Academic Press.

2. Smith, R (2006). "Size of the Moon", Scientific American, 51(78).

That's great, except that numbering has to be done manually, because only <ref>'s know about numbering.

Sorry to be so long, but I'm a newbie who has to choose a referencing style from scratch. That is why I'm trying to understand how the things work. Hugo Herbelin (talk) 11:01, 30 December 2007 (UTC)

---

You seem to understand it fairly well from what I can see above. Perhaps a version of the summary table you produced could be included on the project page. Yes, the problem with trying to combine direct links and footnotes as you've said at the end is that the numbering would have to be done manually, which could become very difficult to maintain as the article develops in the future. Also, something would have to give in the ordering of the citations and full references. The numbered citations as they appear in the text would have to be numbered non-sequentially to correspond to the alphabetized full references (if the full references are numbered sequentially). Either that or the numbers assigned to alphabetized full references would have to be shown out of order.

e.g. say, "The Sun" were written by Edward Young, and "The Moon" by Robert Brown.

It would either have to go...

The Sun is pretty big,^[2] however the Moon is not so big.^[1]

1. Brown, R (2006). "Size of the Moon", Scientific American, 51(78).

2. Young, E (2005). "The Sun", Academic Press.

0r, it would have to go...

The Sun is pretty big,^[1] however the Moon is not so big.^[2]

2. Brown, R (2006). "Size of the Moon", Scientific American, 51(78).

1. Young, E (2005). "The Sun", Academic Press.

--SallyScot (talk) 12:06, 30 December 2007 (UTC)

---

You can use the table for the project page if you want. Maybe extra columns could be added then, like ability to list the references alphabetically. Hugo Herbelin (talk) 13:11, 30 December 2007 (UTC)

The Harvard referencing templates subsection

This subsection leads off with the following paragraph:

Inline author-date citations can be generated in the article text using {{Harvard citation}} templates. Use of the Harvard citation templates can include an automatic link to the full reference, but only if the full reference uses the {{Citation}} template (and the author last names and year match). Links are not generated to full references using other templates or those written freehand.

I object to the "but only if" clause. Consider the the case envisioned by the paragraph as written:

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.{{|Harv|Smith|1943|p=123}}

(and, in endmatter)

• {{Citation |last=Smith |first=John |title=Lorum ipsum explained |year=1943}}

Or, as I've more usually seen in wikipedia articles, something like:

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.<ref>{{Harvnb|Smith|1943|p=123}}</ref>

(and, in endmatter)

===Notes===

<references />)

( this expands as: 1. (Smith 1943, p. 123) — click it )

===References===

• {{Citation |last=Smith |first=John |title=Lorum ipsum explained |year=1943}}

( this expands as *Smith, John (1943), Lorum ipsum explained )

Or something like:

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.{{|Harv|Smith|1943|p=123|Ref=some_made-up_anchor_name}}

(and, in endmatter)

• {{Citation |last=Smith |first=John |title=Lorum ipsum explained |year=1943 |ref=some_made-up_anchor_name}}

Or any of a number of possible wildly differing cases which are linked in a forward direction by wikilinks and in the reverse direction by the browser's "Back" button. Perhaps something like:

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.{{|Harv|Smith|1943|p=123|Ref=some_made-up_anchor_name}}

(and, in endmatter)

• {{anchor|some_made-up_anchor_name}}Y'all oreally ought to read the book "Lorum ipsum explained" by John Smith. Look for it in your local library. -- Boracay Bill (talk) 00:33, 30 December 2007 (UTC)

---

I've changed wording of the subsection to which you objected. It now says...

Use of the Harvard citation templates can include an automatic link to the full reference if the full reference uses the {{Citation}} template with matching criteria such as author last name(s) and year.

However, it could be argued that the link keyword originally was automatic. The alternate constructions you've given above are interesting (if perhaps a little esoteric), but mostly ways of manually overriding the automatic link criteria as far as I can see.

--SallyScot (talk) 12:56, 30 December 2007 (UTC)

Wikilinks from citations to full references

After thinking about it some more (see discussion above), it seems to me that the stuff about {{wikiref}} and {{wikicite}} was all a bit fussier than needs be.

I'd said that all {{wikicite}} is really is simply a 'wrapper' around the full-format reference with made up id matching that of [[#Reference-...]] so that the link between them works.

Well, if you look at the HTML code that {{wikicite}} generates you can see that's quite literally true. It simply puts <cite> tags around whatever is in the reference parameter and sets up cite's id parameter as "Reference-" plus the id value.

So anyway, I've rewritten this part hopefully to help demystify it somewhat.

--SallyScot (talk) 11:24, 1 January 2008 (UTC)

References template?

Gimmetrow reverted two edits I made a little while back (see this reverting edit), which I thought would be ok. I'm realizing now why I thought this was the normal practice - AWB automatically replaces <references/>reflists with {{reflist}} reflists. If this is not the agreed upon practice, why does AWB automatically replace the old system with the new one? TheHYPO (talk) 01:16, 2 January 2008 (UTC)

AWB is a product of some independent group of editors and you'd have to ask them why they chose to add that feature, but it's not based on any consensus that the template is superior. Christopher Parham (talk) 01:28, 2 January 2008 (UTC)

Funny you should ask, but according to Rich Farmbrough, this feature is removed in current AWB builds but not yet in the release version. Gimmetrow 01:59, 2 January 2008 (UTC)

Subscription links in references

I'm curious as to what others think about providing an external link to a subscription site in a reference. There are some editors who believe that if the "New York Times" (for example) provides an abstract to an article published in 2005 (and the reader has to pay or subscribe to read the full text of the article), the link should nevertheless be made. I would argue this runs afoul of External links#Sites requiring registration. Additionally, while a link might provide full-text access today, it won't in two weeks or six months (thus creating a dead link or subscription link). The issue is not addressed in the guidelines, insofar as I can see. Thoughts? Comments? - Tim1965 (talk) 22:47, 3 January 2008 (UTC)

As stated in the lead of the "External links" guideline, "the subject of this guideline is external links that are not citations of article sources." So if the site is used as a reference, and not just an external link, the "External links" guideline just does not apply. I would say that if the subscription site provides URLs that work for a long time, they should be linked to, so those who decided to buy a subscription should take full advantage of it. If the URLs are only valid for a few weeks, maybe it is just as well not to link. --Gerry Ashton (talk) 22:59, 3 January 2008 (UTC)

Most newspapers (such as the "New York Times," "Chicago Tribune," "Los Angeles Times" and "Washington Post, for example) keep articles available for free to the public only for 14-21 days. The URL then changes, and goes subscription. Subscription links exist for forever (insofar as I can tell). Some argue in favor of verifiability, that any John Doe can write a citation. But providing a link provides verifiability for that citation. My argument is "so does looking it up in the public library." A link is only as good as the person verifying that link. Verifiability is no good if the fact being cited is behind the pay-to-view wall. As for paying for the article, Wikipedia provides ISBN links for books, but doesn't have a system for newspaper or scholarly articles. Should authors be trying to recreate that by providing links? I'm doubtful (there's a similar discussion elsewhere on this Talk page). - Tim1965 (talk) 23:17, 3 January 2008 (UTC)

Wikipedia's ISBN system can't be directly compared to newspaper archive sites, because new and used books can be purchased from many vendors at different prices, while the on-line archives of a newspaper are usually only available through one company. Also, verifiability isn't "no good" if the fact is behind a pay-to-view wall, it's just accessible to fewer readers than if the site were free. --Gerry Ashton (talk) 23:33, 3 January 2008 (UTC)

I agree with Gerry; you don't seem to be disputing the validity of the sources, and if we are citing the source, we might as well put a link to the most readily available place to find a copy. I can certainly see how many readers would be helped by this practice (e.g. people who already have free access to such archives) but I can't understand how any readers would be harmed. Christopher Parham (talk) 04:04, 4 January 2008 (UTC)

We routinely use (and encourage) DOIs, and in most cases these link to sites requiring subscriptions. Dragons flight (talk) 05:04, 4 January 2008 (UTC)

References for geographical features

Some of the prose and data on Wikipedia about geographical features is a result of a process that could be called original research: looking at online maps and other changing online services, describing the geographical features visible in them, finding coordinates, comparing the proximity of labels to some location, measuring distances, etc. In my mind this is not a problem, though reliable and verifiable sources would be preferred. What however is a problem is the faithful linking of the services where this work was done as the source of the information, and putting that in the references section. Most of the information has not been published as such in those services, is not static, may be a result of unknown interpolation, and its sources in the services are usually not revealed. For examples, most of the articles with links to mapper.acme.com seem to have references of this type.

Would anyone object if these kind of references were removed (for descriptions of features visible on most modern maps or data derived from non-published services) or converted (to geographical coordinates when referring to a location as a map link)? If not, could these thoughts be reflected in the guidelines in some way? --Para (talk) 19:31, 8 January 2008 (UTC)

You might want to talk to Wikipedia:WikiProject Geography, Wikipedia:WikiProject Maps, or Wikipedia:WikiProject Geographical coordinates. If a coordinate is involved, the latter project's {{coord}} links the coordinate to many map services so there are alternatives if one service changes. -- SEWilco (talk) 20:01, 8 January 2008 (UTC)

Also to elaborate, when the location of a feature is not obvious from the location of the article and the editor has felt it necessary to link to a map, then coordinates should be used instead. The coordinate templates have a source parameter for the same purpose people have used a map link as a reference, but it's rarely used for these "original research" coordinates. In the references sections I have found the following types of links:

• Coordinate reference: "The bridge’s WGCB number is 35-04-18[2], and it is located at 41°45′16.632″N, 80°53′52.584″W (41.75462, 80.89794) <ref>map service link</ref>" (from Mechanicsville Road Covered Bridge)
• Map description: "At this point, the highway runs along a viaduct above state route 92 (Nimitz Highway), passing to the north of Honolulu International Airport. <ref>map service link</ref>" (from Interstate H-1)
• Directions: "Turn left, continue x miles, turn right, continue... etc etc <ref>map service link</ref>" (from Mechanicsville Road Covered Bridge)
• Measurements: "Beginning at the east end (traveling westward), under Canyon Road the tunnel turns SSW (202°)<ref>The tunnel was tracked on TopoZone data on ACME Mapper. The angle was measured using Photoshop. The angles are expressed in conventional navigational cardinal direction values.</ref>" (from Robertson Tunnel)
• Data derived from services: "Elevation 3,618 ft (1,103 m) <ref>Note: Topozone.com maps give an elevation of 3,593 ft (1,095 m), while Peakbagger, and Google Earth give 3,618 ft (1,103 m).</ref>" (from Mount Boardman)
• Data interpreted from a map: "Rocky Mountain, elevation 3,080, is located west of Gaddistown, Georgia, less than two miles west of the boundary between Fannin and Union counties. <ref>map service link</ref>" (from Rocky Mountain (Georgia))

Some of these uses may not be too encyclopedic, but I won't go there and would just like to hear people's opinions on the referencing. Should it be done this way? --Para (talk) 23:08, 8 January 2008 (UTC)

I do work on highway articles in the U.S. Usually, when writing a route description, I just cite distance information, as a route going through a particular town or mountain range should be verifiable on any good map. I usually pull distances from the department of transportation map, citing it with [shameless self-promotion] {{cite map}}. When working on a junction list, which includes mileages for each junction, I prefer using websites with the values readily available. Otherwise, I'll calculate the distances on Google Maps by placing a destination at each junction and adding the numbers up; I always link to the resulting itinerary as part of my references.

I try to avoid using coordinates because your casual reader doesn't really have a frame of reference for relating to those. I know I live around 37 N, but other than that, coordinates don't have much meaning to me, and I find that saying "five miles northeast of Ozark" or "just south of Republic" is more helpful for placing things in my mind. —Scott5114↗ [EXACT CHANGE ONLY] 01:34, 9 January 2008 (UTC)

If I'm reading you right, when something is verifiable on any good map, references are unnecessary. Having a route description in an article would then imply that the information comes from a map.

For distances, it's indeed good to use publications where the result doesn't depend on user input. Many articles however use automatic route finder services as a reference for distances. There's no guarantee that such services always give the same route, or then that the distance will be the same every time. With plotting the points and referencing a link that gives the result, the problem is that the measurements are reproducible only online, only as long as the service uses the same data for calculating a non-linear route between the points, and only as long as that same service is still available. The points used should be given in the article so that the same procedure can be repeated elsewhere, even on printed maps. M62 motorway#Exit list for example gives the coordinates for some of the junctions. There are many tools for visualizing coordinates, and the same information can be included in the "direction format" as well. --Para (talk) 13:30, 9 January 2008 (UTC)

[5] is an example link for Missouri's Route 73. As mentioned above, I try to use these only as a last resort, when I can't find the information from any other source. I know that some editors working in other states like Pennsylvania use Microsoft Streets & Trips or Delorme Street Atlas and then cite the version of the program they used. Then, at least you know that it won't "break" in the way you described.—Scott5114↗ [EXACT CHANGE ONLY] 21:55, 10 January 2008 (UTC)

This is an interesting discussion. If I'm following the conversation properly, then these points might be relevant; if not, hopefully they'll prove interesting in their own right.

• I'm not sure that the act of reading a map or measuring distances on a map should be considered "original research" any more than reading and paraphrasing the type on the page of a published book should be considered original research. Of course, it's important to read a map properly, as it is to properly read (and not misinterpret) text derived from a source of information for the construction of a Wikipedia article. That doesn't always happen, but that is one of the shortcomings of Wikipedia that we must live with and wrestle with.
• Topozone and a number of other popular online map viewing interfaces use (in the United States--I can't speak for other parts of the world) actual maps produced by the United States Geological Survey--considered among the most reliable sources for coordinates and elevation data (in most cases). When you're looking at a location in the USA on Topozone, you are looking at an actual USGS map. Whether or not Topozone actually cites USGS (I believe they do. . .somewhere), that is the source they are using. However, see my next point.
• In the last two decades, the accuracy of coordinate and elevation data has improved significantly. USGS measurements for most United States locations were not derived from modern computer assisted/satellite assisted technology. A number of errors--most of them minor--have been discovered since, and it is likely that many more will follow. So, USGS reflects (in most but not all cases) the best information on hand. . .the best published information that is. You may find that your handheld GPS unit actually offers more accurate data than an official USGS map published in the 1980's. Or not, depending on the quality of the unit and service, and how skilled you are at reading it.
• Many other mapping services (Google Earth) tap directly into satellite information for coordiantes just as hand held GPS devices do. This begs the question: can a handheld GPS device be cited? If so, do you cite the device, or do you cite the satellite network itself? Before you say "No! this would be original research!" Keep in mind that the satellite networks that provide the data to handheld GPS units are probably the same networks that provides data to Google Earth and other commonly accessible applications. Furthermore, when USGS gets around to updating its printed maps, it will also probably rely on the same (or many of the same) satellite networks (hopefully more than one). --Pgagnon999 (talk) 04:27, 9 January 2008 (UTC)

No! :-) Seriously, references must be verifiable. Unless everyone is going to send me their GPS devices (note to self: buy a bigger toybox), the reference (the device) is not verifiable. Maps, on the other hand, can be obtained or consulted. RossPatterson (talk) 04:08, 9 January 2008 (UTC)

I agree with you for the most part, however, to play the devil's advocate: how is the GPS device different from Google Earth software provided the same source is cited via the GPS device, i.e. the sat. network responsible for providing the original data? I use my laptop to obtain the same data you are now reading, but I don't need to mail it to you to cite this talk page.--Pgagnon999 (talk) 04:27, 9 January 2008 (UTC)

A single GPS reading should be similar in quality to Google Earth coordinates in a U.S. city. Google Earth has tried to properly georeference their photos and coordinates in cities should be similar to reality, although problems can be found. Street maps (not images) tend to be generated from data which is strongly linked to geographical locations (such as survey markers). Put a single GPS reading (which varies by 15m/50ft) in Google Earth and you'll see some difference in a city. In rural areas the differences between a GPS location and a Google Earth image is likely to be greater. However, which coordinates are good enough for Wikipedia purposes and how should their source be cited? Is a coordinate considered to be obvious enough to not require a source? As with text, such info might be adjusted by editors with "better" info, so if you're marking something not obvious (such as the original entrance to a fort which was later moved 200 yards) then you need to describe it ("Location of plaque marking original entrance" where the plaque can be seen by someone standing there although not on Google Earth). -- SEWilco (talk) 18:24, 9 January 2008 (UTC)

I focus on verifiability, and that means access to the source and enough info to locate it. Based on that, I'd say that neither a GPS device nor Google Earth are verifiable references, and therefore aren't valid references on WP. The GPS device isn't verifiable because it isn't available to others for verification (unless I get that bigger toybox :-) ) — conceptually, it's the same as an email, unpublished correspondence, or the only extant copy of a book on the Pope's private bookshelf. Google Earth isn't verifiable because it isn't fine-grained enough — it would be similar to citing "Rand McNally" or "Encyclopaedia Britannica". Either of the latter can be made verifiable by narrowing down the breadth of the reference (e.g., "Rand McNally's Big Atlas of Wonkaville" or "Encyclopaedia Britannica Vol. 3 Ch. 25"), so perhaps there's a good way to do that with Google Earth too. RossPatterson (talk) 23:36, 9 January 2008 (UTC)

• As for the actual practice of citing elevation and coordinate data, I think it's no more and no less important than citing anything else edited into an article. In my own work, I've found it most fruitful to check more than one source. This is a pretty good practice in any kind of research. I don't know about the quality of reliability among those internet sources that are vague on where they are getting their info, so it's best to always cross-check to a source that is considered to have longstanding quality standards (such as the USGS). I've found that some packaged software (notably DeLorme) isn't specific enough for very precise measurements, but suffices for general measurements. I've also found that the United States Board on Geographic Names doesn't always associate the name of a landform with the highest/lowest elevation point on that landform (or exact lat./lon), thereby making it a problematic source. Why this is, I'm not sure, but I expect it has something to do with the issue of where the name of the landform is in relation to the contour lines on the map and how the data entry clerk interpreted that relationship when it was entered into the USBGN data bank.--Pgagnon999 (talk) 04:30, 9 January 2008 (UTC)

You're correct about citing references for coordinates. It's just like interesting dates - we cite references for them, especially if they're subject to debate or if various sources disagree. RossPatterson (talk) 04:08, 9 January 2008 (UTC)

Only about 10% of all the 300,000 coordinates on Wikipedia have a source parameter. Some of the rest may have a reference tag, but since they're rarely seen for coordinates, most are probably looked up by people using maps and satellite images, indicating an arbitrary location. When the point has been chosen by a Wikipedia editor, how can there be any reference? I think coordinates should have references (or source parameters, not sure which is preferred) only when there is a static dated publication stating the coordinates for a named point, as otherwise they are not verifiable. --Para (talk) 13:36, 9 January 2008 (UTC)

I'm not sure I'm following you. By "source parameter" are you referring to, let's say, cooordinates linked to the GeoHack? Do map and sat. images produce "arbitrary locations?" I'm not sure that's so; if it is, it may be helpful to determine which ones aren't reliable and which are. As for a "point being chosen by a Wikipedia editor", well, wny not? If maps are sources of information, then reading a map (and by extension, doing simple math to specify that meaning) is not different than reading a book and distilling meaning from it, then translating (or paraphrasing) that information into a Wiki article. As for "static dated publication" I'm not sure that such sources are any more accurate than some of the online sources you question (see my example of the USGS above). Furthermore, how exact should map translations be? When I measure the distance between two cities, for instance, it isn't necessary to be accurate to more than a rounded mile/km, less so if the distance is great or the routes between the two points vary--the kind of accuracy you'd find on just about any online source, and the kind of accuracy you'd expect from anyone reading a map and making no-brainer measurements on it with a ruler. I think we have to assume "intelligent intent" with regard to editors and their capabilities to read simple maps and draw simple measurements from them. --Pgagnon999 (talk) 14:13, 9 January 2008 (UTC)

The style manual instructs coordinates to be entered using a coordinate template, so we can assume that all coordinates people have worked on are in that common format. Since all such templates use the common parameters available with coordinates, the source parameter is also available.

In my opinion maps and satellite images do not indicate the location of large scale features such as mountains with a precision necessary for the 1ft elevation precision that many articles use. If the coordinates of the article are appropriately rounded for the scale used, and there are no coordinates for the highest point, how can anyone verify the elevation if the source is just a map or a geographical service with similar information? If the coordinates are an estimate by some guidelines, be that the center of the feature, highest point, some politically chosen point, or just rounded to the closest unit appropriate, and the coordinates can be verified in other maps or services to show the same general location, then the source of the coordinates is all of them together and there is no need for references.

There's no problem if the reference is to a site that always gives the same data (ie. allows linking to data from a given date), that's the case with books, lists, and some online services. Accuracy aside, they are verifiable. Many online services however are not; the elevation mentioned in the Mount Boardman article for example doesn't seem to be shown as such in any of the referenced sources for those coordinates, which makes the references useless.

Perhaps instead of references for those simple measurements and data from common sources, there should be a tag indicating that the data has been verified at the time of the edit, and that all the geographical data in the article is available from the services linked through the coordinates? --Para (talk) 17:56, 9 January 2008 (UTC)

Sorry, but I still don't know if I'm clear about what you are trying to say. I looked at the Mount Boardman article; clearly if you compare the elevation data in the article with the USGS topographic maps (Via Topozone, via GeoHack), you end up with a USGS topographic map showing the mountain but indicating an elevation higher than indicated in the article. So, someone is wrong. The problem here seems to be either a typo/mistake, or the fact that the article writer used a different source than indicated in his reference. But this is an editing problem; it can be corrected by re-editing the article and supplying referenced elevation data--via whatever source. Of course, the writer never specified which of ghe GeoHack sources he supposedly used. . .is that what you are getting at? Yes, it would be better for an editor to be specific about which one he used. . .if he used any of them at all. As for the "1 foot precision" in citing elevations, I think it's generally accepted that such measurements, no matter what the source, are an approximation, not an absolute--even for elevations listed as precise on USGS maps. As for estimating elevation where "exact" elevation is not given, it's obviously important to let the reader know that the measurement is an estimate, better yet, indicate the margin of error inherent in the estimate in a footnote. For instance, "Estimated at 500 feet +9 feet/-0 feet" for a summit with a highest contour of 500 feet and a coutour interval of 10 feet. I'm not sure if that addressed your comment. .. --Pgagnon999 (talk) 20:32, 9 January 2008 (UTC)

When the data entered to Wikipedia is not from any single source, but is averaged, estimated or outdated values from an undated source, is not visible as such in the services, and is not a result of a simple calculation (to the reader, as he doesn't know the source numbers but only the result), it is not the same as ideas distilled from words in a book. It is changing the information found from the source, and you can no longer say that the composite information is from that source, but from "all the available topographic services on the map sources page". This is not an isolated problem in a couple of articles, but is evident in all articles that cite services with geographical data without a date or source, and is even worse when "simple measurements" have been done. --Para (talk) 21:21, 9 January 2008 (UTC)

Again, and I'm not trying to be difficult, but I'm not entirely sure what you are trying to say. I think I'm closer to understanding you, but not quite there yet. What specific "undated" sources are you referring to? Are you indicating that the problem is with the fact that some editors cite Geohack itself as a reference for geographic data instead of verifying the actual source data via the options listed in Geohack? Or are you indicating that the actual possible sources listed in GeoHack shouldn't be cited and are "undated"?--Pgagnon999 (talk) 21:34, 9 January 2008 (UTC)

I'm saying that when data is not clearly from any single source where it can be verified, it should not be referenced as if it was from such a source. If the data is originally from a reliable source, but the reference isn't or can't be made to mention the original source and date of the data, the reference should not be made at all, as it is of no use. Google Earth and NASA World Wind for example can easily be used to get some elevation number to make articles look pretty, and other editors may even be able to find the same number from that source for some undefined period of time after the addition, but there is not enough information to make a complete reference for people to verify later. With the Mount Boardman article for example, where we now have to guess if the sources have changed or if the editor made a mistake, had the reference been made with identification and date of the original dataset, we would have the necessary information to find out if the source data has indeed changed since the edit or not. All references to services with no identification information of the source of their data, where the same information either can't be found at all or can be verified in any service with similar data, should be removed from Wikipedia. --Para (talk) 00:33, 10 January 2008 (UTC)

From purusing your talk page, it seems that you have a strong geo-tech background. Although I certainly respect that, I would encourage you to keep in mind that, since you've brought this issue to a talk page that is not specific to geo-technology, it would be helpful (to me anyway) if you could use specific examples of what exactly you are proposing, and how exactly you see Wikipedia changing. When you say "original dataset" specifically what dataset(s) are you referring to? Which sources specifically are you suggesting that we do away with? How specifically do you envision they might be done away with? How would that be implemented? How do you see it happening that editors can be compelled to stop citing, for instance, Google Earth, and start citing these "datasets" you mention? Seems like a tall order. I'm wondering, if I'm reading you properly (and I still don't know, as your language is still vague or maybe I'm just thick), if perhaps you are suggesting something a little extreme here. There are clearly instances were absolute data should probably be cited (for instance with regard to the finer details of nuclear physics), but in other cases, I do believe that a Rand McNally road map is a legitimate enough source, let's say, to get a general measurement in miles between two cities, regardless if Rand Mcnally reveals its original data source or not. In other words, I don't think it helps Wikipedia to insist on such rigor across the board (if indeed that is where you are coming from). . .not to mention the nightmare of trying to enforce it/make it happen. If that is what you are suggesting, I'd have to strongly oppose such a change. --Pgagnon999 (talk) 02:42, 10 January 2008 (UTC)

It's not specific to geo-technology that full references should be given when citing a source, to allow people access to the same data at any later date. That's why static publications should be preferred as sources. When you can't find a reliable static source for something, instead of citing an everchanging online service, you should go to the original source or not give any reference at all. For example, USGS datasets with sufficient identification can be found from the Geographic Names Information System and National Elevation Dataset among others. Many of the services people reference on Wikipedia may use the same data as a source, but if the service doesn't mention where and when the data is from, and that information is not carried over here to the reference, it might as well have been made up as far as reliability and verifiability is concerned, making the reference worthless.

I'll go to specific details in a #Georeferencing recommendations section then, when I'm up to writing one, using the examples from above. Getting rid of such bogus unverifiable references wouldn't be harder than anything else Wikipedia instructs editors to do. It doesn't need to happen overnight, but there can be guidelines here and in related wikiprojects discouraging bogus referencing, vigilant editors zapping such references on sight, informing mistaken editors on proper referencing, etc. Note that I'm only talking about geographical information here, and of sources that can change, or of information that is the same everywhere. --Para (talk) 14:00, 10 January 2008 (UTC)

Thanks for working to be clearer :). Although I agree with your statement that "bogus" references to geographic data should be discouraged, I disagree with a unilateral labeling of all sources that do not cite specific "datasets" as "bogus" under all circumstances. For instance, Google Earth, regardless of where it is getting its data, provides direct, easily accessible visual data about the surface of the earth. Websites change all the time, and that is why it's important to include a "Retrieved on (date) reference." Let me give you a specific example. for instance, let's say you wanted to inform a reader about the (relatively) current dimensions of an ongoing stone quarry on a specific mountain--let's call it North Peak. As the article is about North Peak, not the quarry, it's important to a) prove that the quarry exists; and b) provide an estimate of the size of the quarry--a general, not absolute--measurement which is going to be "dated" no matter where the information comes from, unless there is someone out there at the quarry measuring its expansion every time a chunk of rock is removed. So, for the article about North Peak, one might cite a measurement of the quarry taken from Google Earth, and in the reference section, note when the information was retrieved from that source. Of course, you can argue that it would be better to go to the source of the Google Earth data instead, and you're right--it would be better but it wouldn't be necessary. And, as I've argued elsewhere in this article, and have provided specific examples, the Geographic Names Information System is not necessarily more reliable than some of the sources you suggest that we shun, and in many cases it is less reliabile. And yes, I do believe that reliability is something that should be important to us. If I have a source of information that provides consistantly more reliable data than another "official" source of data, then I'd be a fool not to go with the more reliable source, regardless if or not they display their "original datasets" as long as they have a reputation of providing accurate information as it applies to what I am working on, and the level of detail I am working with. I would, however, have no problem with you or anyone editing articles, on an individual basis by replacing a less precise source with a more precise source, but again, I think this has to be done on a case by case basis; I think it's determiental to force it upon all articles. For instance, if it was decided that the USBGN data lookup feature should be the source of elevation data for mountain summits in the United States, we'd end up with quite a mess of misinformation on our hands, as it is not (consistently enough) a reliable source for such information (see my comments and examples below--if you can find them in this anaconda), regarless of the fact that it can cite its original datasets. My fear (and maybe it is unfounded) is that you are suggesting a unilateral change to Wikipedia and a universal branding of certain sources (you have yet to name specific ones) as unsuitable as references because they do not reference the absolute sources of data they are working with. This seems a little extreme to me. Aside: Did I also mention that USGS datasets are often decades more dated (and therefore dependent on less reliable technology) than other current sources? --Pgagnon999 (talk) 16:03, 10 January 2008 (UTC)

• Finally, I should point out that it's important to match up the proper coordinate format when linking from an article to the Mapsources GeoHack Wikipedia tool; if not, you'll end up directing your readers to Mars.

--Pgagnon999 (talk) 03:50, 9 January 2008 (UTC)

The United States Board on Geographic Names is probably often used for coordinate information although the source is often not cited. Unsourced coordinates often have the USBGN numbers, although it is possible the numbers came from another source (or from the USBGN through another source). However, reading a GPS device should be considered as an acceptable method; the USGS uses volunteers with GPS units to collect coordinate information. If we want to define a coordinate manual of style (driveway, main entrance, or center point?) we'll do that as we have for other source materials. Trivial calculations, such as taking GPS readings on opposing sides of a building and using halfway between the two locations as the location of the building, should also be as acceptable as paraphrasing text is. Another editor with a more precise or authoritative source/device may later correct the info. The source should be specified if it is not obvious or considered common knowledge. -- SEWilco (talk) 17:08, 9 January 2008 (UTC)

With regard to your comment "The United States Board on Geographic Names is probably often used for coordinate information", I'd like to suggest that the USBGN online data lookup feature is probably less reliable than a handheld GPS device for a number of reasons; see above and also related discussion at Wikipedia talk:WikiProject Mountains#peak lists?.--Pgagnon999 (talk) 17:21, 9 January 2008 (UTC)

Yes, I think the USBGN is authoritative and official, but probably suffers many of the problems one would expect from a government-affiliated official decision making organization. The USBGN FAQ [6] 29 does state that updates are frequently applied, whatever those are. The USBGN procedures are focused on names and I suspect that USGS workers update coordinate info but I haven't found details. Whether USBGN numbers are correct for Wikipedia use (such as city coordinates being of oldest location, such as city hall) depends upon what Wikipedia wants. At a minimum, USBGN is a citable source of some authority. Locations which are not in USBGN require other sources. -- SEWilco (talk) 18:04, 9 January 2008 (UTC)

Here's a specific example of the type of problem one is likely to encounter when using the USBGN: Peak Mountain. If you refer to a USGS topographic map of the mountain (such as this one duplicated on Topozone) you'll note that A) The words "PEAK MOUNTAIN" are shown along the entire length of the mountain ridge, and B), According to the contour lines, the summit is over 700 feet. However, if you look it up via the USBGN (see here), you'll see that the summit is listed as 581 feet. If you then click on the feature name (Peak Mountain) in the USBGN lookup results, it will give you the option to view the mountain via Topozone. Click on that, and you end up with this, which shows you the southern ledge of Peak Mountain at 672 feet, not 581 feet, not the actual 700+feet summit. Obvioiusly, this throws off coordinates and elevation. If this were a rare or isolated incident, not big deal, but it's not. I can cite several more such inconsistancies all within a radius of less than 30 miles. If that is representative of the degree of reliability on a national level (and why wouldn't it be?), I think it brings into question any reference using the USBGN alone. --Pgagnon999 (talk) 18:20, 9 January 2008 (UTC)

USBGN FAQ [7] 16. # How accurate is the elevation data in the GNIS Database? How was it measured? -- SEWilco (talk) 18:34, 9 January 2008 (UTC)

The USBGN is one source which can be cited on Wikipedia. Other sources maybe be better for specific locations or uses; for a mountain the altitude used by some mountaineering record tracking group might be suitable. If you can park a precision GPS unit on a point for 24 hours that might be better (location of the first Sears store?). -- SEWilco (talk) 18:34, 9 January 2008 (UTC)

tl;dr the above, so this might have been covered, but on U.S. roads I generally cite Acme Mapper, which is a pretty nice Google Maps/USGS topo combination. For instance, see the route description of California State Route 190. If there's something that's not obvious from any of the three types of imagery - road, satellite, or topo - then I cite a better source. For elevations, unless the topo has a number right at the point, I give an approximate figure based on the contour lines. --NE2 12:41, 10 January 2008 (UTC)

Ok, so your preferred map service or user interface is ACME Mapper, but is that tidbit suitable for articles? Any map from after the geographical feature was built provides a similar view that you can look at when writing a description. If something is obvious in the imagery you have looked at, and others, and its location is noted in the article, is it really necessary to cite any particular source, when the same information is available in so many other services for the same location, without any interpretation or other processing necessary? --Para (talk) 14:00, 10 January 2008 (UTC)

Not all maps are equal. One might not be as up-to-date, while another might show a different name for a road. Citing exactly where I got the information lets readers check the reference. --NE2 15:08, 10 January 2008 (UTC)

You recognise that the date information is important, but you're still referencing something that has neither date nor source. That's the same as citing a friend of a friend who read the information from a reliable source; the information will not be verifiable, since the source you used is known to be dynamic and so it may not have the same data from the original source anymore, or may not be available at all. Shouldn't you also mention the discrepancy of the sources in the article and cite the different map in addition to the maps that have common information, instead of making your own judgement on which source is right and citing an aggregate service where you can't know where the labels will be tomorrow? --Para (talk) 15:46, 10 January 2008 (UTC)

That's why I say when I accessed it. I'm not saying that there is a discrepancy, just that if there is one it will be clear that there is one. --NE2 16:51, 10 January 2008 (UTC)

But it won't be clear, that's the point. Look for example at the articles Google Accused Of 'Airbrushing' Katrina, Google's View of D.C. Melds New and Sharp, Old and Fuzzy, and Censorshopping around the Netherlands. Google Maps is such an everchanging cocktail of data that you can't guarantee that others would get the same information tomorrow, and they have no way of knowing the data has changed from since you looked at it, since it's given without sources as if it was factual information. If however you're writing about some large scale present day topic such as roads, small differences in the data are insignificant and it doesn't matter whose map you're looking at. What information does a reference to a single service then convey and will it help convince readers of something? --Para (talk) 01:10, 11 January 2008 (UTC)

I understand your frustration with the changing nature of data provided by internet geographic services, but isn't that the nature of the internet as a whole? I think that's fairly well understood by anyone that when you read an article that references by an internet source, that source may change at any time, without notice or indication that it has done so. But that doesn't mean the source isn't worthwhile and shouldn't be cited. And sure, Google may have airbrushed Hurricane Katrina, and there certainly have been inaccurate news articles that have been printed by generally reliable sources such as the AP, BBC, whatever, but occasional incidents of inaccuracy or fraud should not mean that the baby should be thrown out with the bathwater. --Pgagnon999 (talk) 01:22, 11 January 2008 (UTC)

To jump in here, as it coincides with another sub-discussion going on inside this beast, I agree with NE2, and I wonder if you are insisting on too rigorous a standard. I think that there are certainly instances where such rigor is justified, but I don't think it is necessary to enforce it in every instance. A map published by a reputable publisher should be citable, regardless if it reveals its original data set or not. Of course, it can be challenged as a source, and replaced with a better source, but it should not be automatically regarded as unsuitable--Pgagnon999 (talk) 16:17, 10 January 2008 (UTC)

I don't see what the problem is here. Encyclopedias do not contain page after page after page of stubs about irrelevant hamlets because it would obviously detract from the important stuff, since if people wanted to know how to skin a cat or find the mystical land of Toul Kork, they would go somewhere else. However, Wikipedia is not paper, therefore it may be spammed by bots injecting directory lists that could easily be found on maps.google.com and Google Earth. For an idea of what I'm talking about, simply refresh Special:Random over and over. It is important that we include this stuff, because someday, the ghosts of the Native Americans that died in Adobe Creek might rise from the dead and want to contribute to that article to tell us a bit more about it. Because obviously, the squirrels and bunny rabbits there aren't talking. Until scientists manage to invent a means of immortality, that's not going to happen of course, and it's just going to be a worthless stub -- but give it some time. Consensus is inescapable and will fix things, eventually. Zenwhat (talk) 22:28, 12 January 2008 (UTC)

I must admit to a bit of confusion with the Wiki policy upon citing sources for geographical places altogether. I have been happily adding latitude and longitude for a few years now. This involves "improving" coordinates of existing articles (some Wiki coords come from geonames where previously "near enough was good enough" - I'd find that geonames sometimes locates a place in some nearby field rather than in the centre of an urban area) and also sourcing brand new articles. People leave you alone of you are redoing the coordinates. If you are creating a new article for a village, people now want contributors to cite sources "otherwise an article may be deleted". I think this is overkill - I added Morefield last month as an example. My POV:

• The Wikipedia is infinitely expandable. At least one person comes from every village. I'm not sure waiting until somewhere becomes significant for some reason matters. If it's on the map, I'd want it in the Wikipedia. I leave whether every hillock needs cataloguing for others - I'm only interested in human settlement.
• Indeed sometimes there is no true source. Yes I may have located it using Google Maps or Virtual Earth. But those sources did not suddenly spring into existence this decade - they are built on the work of predecesors. Originally indeed somebody 100 years ago went around with surveying tools which became the Ordnance Survey which became a source for Multimap. Personally my source is "well there it is on the map I'm using".

Does citing sources for geography need a rethink generally? --Scotthatton (talk) 18:32, 14 January 2008 (UTC)

References are good, but only if they lead the reader to the information cited. Though anything on the internet can change, most reliable online sources cited on Wikipedia are publications that aren't expected to change. If they disappear, it is often still possible to find the material again, and they usually have an author who can be contacted. With dynamic databases such as most online map services none of this is true: their very nature is that they change (or in people's minds, improve), there is no way to see what a service's imagery at a certain date in the past looked like, there is no identification in the material, and no way to contact the authors. Such references are therefore of no use, and only give the false impression that the article is well referenced, when on further inspection the references are not verifiable and all we have is the editor's word for it. To compare with other topics, it is not customary on Wikipedia articles about companies or currency to cite current stock exchange or exchange rate services. With such data there is no expected truth the services would be trying to approach and Wikipedia to catch up on, but only different periodic values, which is why those articles cite published historical data. When there is a truth to something we write about, it shouldn't be an editor choice to decide whose data is closest to it, but it seems that that's exactly the practice in Wikipedia. A rethink is certainly needed. --Para (talk) 12:10, 16 January 2008 (UTC)

Documentaries and Interviews

Just one question, how would one cite a TV documentary or interview? A lot of the time these things don't appear on the internet or on a news broadcast, so it's difficult to include information from a source like that in an article. BalkanFever 04:29, 9 January 2008 (UTC)

This sounds like a job for {{cite video}}. RossPatterson (talk) 23:17, 9 January 2008 (UTC)

---

Here's a real example using {{cite video}}...

• Sherratt, Dr Andrew (Presenter); Sarah Marris (Producer); with Daniel Seibert, Dr Françoise Barbira-Freedman, Dr Tim Kendell, Dr Jon Robbins and Sean Thomas (1998). Sacred Weeds: Salvia divinorum (video) (Documentary). UK: TVF Productions (for Channel 4). Retrieved 2007-08-08.

--SallyScot (talk) 11:32, 12 January 2008 (UTC)

Referencing different pages in the same book?

I have been learning how to add citations and references for a while now, but I still do not know exactly how to make multiple references to different pages in the same book. I could just reference a book without a page number, give the reference a name and make multiple references but I want to make my citations more precise.

I hate not backing up statements and claims without the exact page number listed. Despite checking many times, over a period of months, I find that the help pages are not clear enough for me to understand how to do this. - Shiftchange (talk) 22:51, 11 January 2008 (UTC)

One solution is to use {{Harvnb}} and {{Citation}} in combination with one another. See Philippine-American_war#Notes for an example. Click wikilinks to get from footnotes containing (possibly page-numbered) references to the associated citation or a work to which the references apply. Use the browser's Back button to get back. -- Boracay Bill (talk) 00:55, 12 January 2008 (UTC)

The inline citation method Short footnote citations with full references covers this. It's documented with examples on the project page. --SallyScot (talk) 11:38, 12 January 2008 (UTC)

I think the way most people do it, Shiftchange, is simply to write <ref name=Smith25>Smith, John. ''How to Write References''. Random Publisher, 2008, p. 25.</ref> Then if you reference the same page again, write <ref name=Smith25/>. This avoids the need for citation templates, which can make the text hard to edit. SlimVirgin ^{(talk)(contribs)} 13:37, 14 January 2008 (UTC)

Scrolling lists

I notice that this was recently added:

"Scrolling lists, for example of references, should never be used because of issues with readability, accessibility, printing, and site mirroring. Additionally, it cannot be guaranteed that such lists will display properly in all web browsers."

Does anyone know where it was decided that it's best not to use these? I've seen them used in articles with very long Notes or References section to great effect. I can see that printing would be problematic, but then lots of issues in articles can cause problems with printing, and we're an online project, not a paper one. Apart from that, though, are there real problems with these scrolling lists that I'm not aware of? SlimVirgin ^{(talk)(contribs)} 10:22, 14 January 2008 (UTC)

I have had no involvement here either but I can see Wikipedia_talk:Cite_sources/archive18#Scrolling_Reference_List, Wikipedia_talk:Cite_sources/archive19#Scrolling_reference_lists, Wikipedia_talk:Footnotes/Archive_7#scroll_box_for_references and Wikipedia:Templates_for_deletion/Log/2007_June_11#Template:Scrollref. To me such lists look to be a bad thing. Thincat (talk) 10:47, 14 January 2008 (UTC)

As it happens, I don't think scrolling lists should be used either, particularly for References, which would be at the bottom of the page, where you'd be using the browser's own scroll bar anyway. However, I don't really agree with the inclusion of the argument that it cannot be guaranteed such list will display in all web browsers. Arguments made against scrolling lists should legitimate. Concerns over readability and printing still stand, but the argument of browser compatibility is being overstated. --SallyScot (talk) 11:23, 14 January 2008 (UTC)

I'm wondering what it is about scrolling lists in particular that is problematic. Image size and placement, for example, will look different on different browsers, will sometimes not print out well etc, but we don't disallow images for that reason. SlimVirgin ^{(talk)(contribs)} 13:31, 14 January 2008 (UTC)

If I recall correctly, most mirror sites can't handle the scrolling lists; but that may have changed since the last time I checked. Kirill 13:45, 14 January 2008 (UTC)

The other question is what benefit they provide. There is no particular reason to save vertical screen space within a page when most browsers already have a scroll bar on the right hand side of the window →. In most of the cases I see where the scroll box is desired, a better solution would be more judicious use of references, a change to a referencing format that uses less space, or a combination of the two. — Carl (CBM · talk) 15:34, 14 January 2008 (UTC)

Style guideline for PD sourced content

Template:RFCstyle Does WP:REF support that all (all, as in everything without exception) PD-sourced material be placed in quotes to avoid the appearance of plagiarism? If PD-sourced material is not in quotes, is it proper to remove the offending text? Please say it ain't so! -- Paleorthid (talk) 19:18, 16 January 2008 (UTC)

The same editor removed a map and updates which interfered with his wrapping PD text as quotations. He places imaginary plagiarism restrictions above that to promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries. Or, in this case, this wasn't even expired copyrighted material, as the work was created in the public domain. -- SEWilco (talk) 20:39, 16 January 2008 (UTC)

SEWilco, that is misleading. You know I made effort to save your contributions here: Talk:Bathhouse Row#Alternative sources, for editing back into the article when more peace prevails. To others, please see the history of edits of Bathhouse Row, reflecting many combative reverts, and the talk page, if you want to review SEWilco and my roles. doncram (talk) 21:12, 16 January 2008 (UTC)

The PD text was still adjacent, with cosmetic blank lines (proposed by another editor), although better organized than in the original, with the source cited at the end of each section. There was no need to destroy the improved presentation and remove material. And you didn't bother restoring the new material, leaving the article for readers and editors with older text and large blocks of text protected from alteration within quotation markings. You say below that "a huge block of text" is intimidating, yet that is exactly what you insist upon. -- SEWilco (talk) 21:21, 16 January 2008 (UTC)

To give some more info to Paleorthid's question, I have indeed added the "nofootnotes" tag to a number of articles covered in "Category:Wikipedia articles incorporating text from public domain works of the United States Government". For some of these articles which show some effort to properly reference other sources, I have added a note to the talk page such as this one related to one of my edits that Paleorthid questions. In response to a a similar posting I made on Talk page of "2007 Brooklyn tornado", an author/editor expressed appreciation for my calling attention to the appearance that the article was poorly sourced due to the use of that "USGovernment" template, and has chosen to fix the article so as to be able to remove that disclaimer. In other cases I have just added the "nofootnotes" tag. In some cases I removed offending text (leaving the appropriate external link to the text source). In one case I returned to add more sources and to set aside copied text in a block quote, see this diff on James R. Allen article. I think all of these interventions are helpful and justified. doncram (talk) 21:01, 16 January 2008 (UTC)

I have stated that these types of edits on articles involving big hunks of copied text are needed to avoid the appearance of plagiarism, but that is only one reason. Another reason is to clarify that, in some cases, virtually all the text is copied, perhaps slavishly, from a source that perhaps should be questioned. I noted in the Talk:James R. Allen case that the official U.S. Air Force biography which was the source, had neglected to discuss potential controversy about his role as superintendent of the Air Force Academy when women were admitted, which later led to great controversy and scandals. I don't know in that case whether Allen was involved constructively, obstructively, or not at all, but I do question whether the USAF official biographies are selective in what they report in ways that excludes anything that might reflect poorly on the USAF. I think it is highly appropriate to segregate by quotations or block quotes anything that is directly copied, to avoid being slavish dupes of dated or otherwise questionable sources. doncram (talk) 21:01, 16 January 2008 (UTC)

Another reason is to facilitate entry into the editing of an article by other editors. When there is a huge block of text copied in from another site, it is naturally intimidating to many potential editors. To begin to revise or add to an already big article that may be stated very authoritatively, is difficult, unless and until one understands that all of the article, or huge chunks, is merely copied from one source. This came up in somewhat heated discussion with SEWilco of Bathhouse Row where copied text is currently segregated in block quotes; it certainly applies to Duquesne Spy Ring which is mostly 33 profiles and photos copied from here. doncram (talk) 21:01, 16 January 2008 (UTC)

Doncram says that a huge block of text is intimidating, yet that is exactly the format he has forced upon Bathhouse Row. The copied text was "segregated" by him, replacing a better organized and well sourced format with updates which he removed. -- SEWilco (talk) 21:21, 16 January 2008 (UTC)

Copied text seems unencyclopedic. Shouldn't an encyclopedia include condensed discussion, that is be shorter than the sources that it cites? It does not add value to merely copy. It potentially adds value to call attention to material, by selective quoting and citing, or by including an external link. doncram (talk) 21:01, 16 January 2008 (UTC)

I do appreciate Paleorthid's raising this question here, which may be an appropriate place to raise it. I look forward to your comments. I have also made a point to raise the issue in WikiProjects involved in many of the articles, as here:Wikipedia talk:WikiProject Military history#Copied material in Military History articles, and quality ratings and here Wikipedia talk:WikiProject Aviation#Copied material, inappropriately sourced, in aviation articles. doncram (talk) 21:01, 16 January 2008 (UTC)

As I told Doncram (who seems to have then ignored me and gotten into the same dispute with other people) PD text doesn't need to be in quotes. There's no legal or ethical requirement for this... even our copyright/plagiarism critics have never complained once about our use of PD text. There are much better issues to fret over. --W.marsh 21:30, 16 January 2008 (UTC)

• No - per WP:CREEP first and foremost, but among other things per above. MilesAgain (talk) 22:09, 16 January 2008 (UTC)