At UMass, we've been using a blog -- a catablog -- as one means of providing
access to our general manuscript collections, with links out to EAD files
and other resources (http://www.library.umass.edu/spcoll/umarmot/).  Users
can supply comments and add information to descriptions or pages, all
moderated. Thus far, the comments we've received have been restricted to
spam, more spam, reference questions, spam, and pats on the back. This may
evolve as the site evolves -- it has been up for only a brief time and is
still "in development" -- but perhaps because of the nature of our
collections, the comments have not been uniformly valuable.

Personally, I'm not wild about a wiki approach to finding aids for some of
the same reasons already mentioned, but even more because of the sociology
of wiki writing.  However "accurate" a wiki article may be, the result too
often tends to be a watered down, pale imitation, stripped of meaningful
analysis and acceptable to a large number only because it avoids any real
issues in interpretation or historical context.  When I want to know a birth
or death date it may be fine, but when I'm wearing my historian hat, I'm
usually far more interested in the interpretative layers than the so-called
factual layers (which, as has already been pointed out, have their own
problems).  Perhaps in some contexts a wiki-aid might work well, but
dinosaurs like myself go to some lengths -- not always successfully -- to
craft an image of who we are and what we do, and for me, the wiki approach
doesn't quite get us to the end of the Cretaceous. 

Sorry for the metaphor.

--R

___________________________________
  Robert S. Cox
  Head, Special Collections
          & University Archives
  Adj. Prof. of History
  W.E.B. Du Bois Library, 154 Hicks Way
  University of Massachusetts, Amherst
  Amherst, MA    01003-9275
       (413) 545-6842
___________________________________