Wikisource:WikiProject DNB/WP referencing

This is a page about general DNB referencing on Wikipedia, and the effort of this project to supply it. The problem is protean: there will always be a large number of ways that editors use to reference the DNB, and correspondingly a need to search for such references efficiently.

Multiple easy automated ways to search for DNB references on Wikipedia, that each turn up short lists, are a desirable outcome, in fact: much more so than trying to generate a "master list" that takes the form of a huge, undifferentiated backlog and that needs constant updating. Fuller details are being posted in userspace, to try to keep this page comprehensible. Please post detailed comments about instances where the DNB is referenced in other ways on the Talk page of this page.

Objectives
It hardly needs saying that one aim of the project is to support DNB referencing by direct linking to Wikisource, as the best solution. To understand the scope of that ambition, namely to replace other ways of referencing by the use (mainly) of w:Template:DNB Cite, we require some preliminary understanding


 * 1) of how such referencing is carried out, and
 * 2) methods of finding the WP articles needing direct references.

The basic template for use of DNB text
There is a clearcut and ongoing effort at /Data capture to supply links to Wikisource's DNB articles for every use on Wikipedia of the template w:Template:DNB that should indicate usage of text from the 1885-1900 DNB. This is just a small but relatively tidy corner of the much wider issue of tracking DNB references on WP.

Examples

 * References to Google Books versions of the DNB

This is particularly vexing, though not apparently the biggest issue (maybe a few hundred instances?). It is annoying because whether or not you can read the Google Books version may depend on where you are in the world, meaning that these references may have been added in good faith by editors not realizing that they are creating a problem for some readers. There is a complication of multiple editions available on Google Books.


 * References to archive.org

Some editors currently reference the DNB by linking directly to the versions online at archive.org. Where this is to the "read online" paginated version, that is not too bad for the reader. It would still be an improvement for editors to be able to copy in text to the article to expand it. For example this search on a pattern from the "Toronto" scans shows dozens of article hits.


 * References to the ODNB subscription site

These run into several thousands. Clearly placing a DNB reference alongside the ODNB link will help most readers (use w:Template:DNBfirst).


 * Berkshire History

At http://www.berkshirehistory.com/. This is a smaller problem, where the website contains numerous edited versions of DNB text. From the point of view of referencing, the original DNB article is going to be superior.

Methods
The useful general tool is http://en.wikipedia.org/wiki/Special:LinkSearch. It finds external links containing a given pattern.

For example this link picks up the list of links to the Berkshire History biographies.

The pattern searches may need localisation:


 * For vol. 28 is a pattern for books.google.co.uk
 * For vol. 31 is a pattern for books.google.com

Conventional search
"Dictionary of National Biography" as search term simply throws up a very large number of hits; and as more DNB material is added there will be more hits, not fewer, since some of the hits come from referencing templates. More discriminating searches are needed.

One type is based on ISBNs for reprints of the DNB. For example this search for vol. 47, Adamant Media reprint. It overlaps with, but doesn't duplicate, an external link check for the Google Books version.

A search for a supplement.

Compilation
To develop our “most wanted” listings, we need to compile lists of articles requiring DNB references, with names for the biographies (where those are different), and then check against existing articles to track how much has been implemented. This would be the logical outcome of a big drive to search first, create articles later. On the other hand, "most wanted" as a concept needs some work.