Natural history museum collection databases typically contain spatial
coordinates (latitude-longitude or UTM) for less than five percent of
their specimens. Instead, most specimen localities, especially
historical data, are in textual format, i.e., "2 mi S of Bakersfield."
Without spatial coordinates these data are effectively unavailable to
GIS analysis or even simple mapping.

Due to the growing use of GIS applications, there is an effort
underway to make these collection data available to spatial analyses. 
I am interested in compiling examples of as many different
retrospective georeferencing methods as possible, specifically:

 -- tools for converting textual localities to coordinates
 -- current efforts by natural history museums and other biological
collections to georeference their historical localities
 -- thoughts and theory on the problems inherent in converting a
textual, sometimes vague, place description into a point, line or polygon

I am aware of several efforts and will be using the following:
 -- http://www.calacademy.org/research/informatics/GeoRef/index.html
(I am part of this project)
 -- the MANIS project
 -- discussions on Taxacom listserv
 -- "heritage" programs such as the CA Natural Diversity Database
 -- Texas Tech University http://www.nsrl.ttu.edu/projects/GIS/GIS.htm*

*Semi-academic references such as this TTU link are just fine as are
anecdotal or nonscholarly references, although journal articles in
geographic, cartographic or biological journals are slightly more
valuable in the long run.

Thanks!