My experience as a cataloguer was that we could not establish personal
author names uniquely and correctly without a lot of bibliographic
research. As a special library with a mission to provide health
information, we decided that this was not a good use of our time. We
reduced the amount of effort put into descriptive cataloguing and focused
on subject analysis. This was back in about 1980.
AACR is a comprehensive standard for descriptive cataloguing. Together with
MARC and MARC-based systems, it ensures that searches for known authors or
known resources can usually be successful. The situation for subject
cataloguing is a very poor second - this can be demonstrated if you try to
do a complex subject search on a library catalogue, compared with searching
on a database like Medline.
If Dublin Core metadata is concerned with resource discovery, then we must
not neglect subject areas. This includes working with search engine
designers.
And we have to continue to keep in mind: who will create metadata for a
resource? how much effort will they be prepared to put into it? who will
pay for the time involved? is this a cost-effective use of the information
budget?
Prue
Prue Deacon
Web Business Solutions
Commonwealth Department of Health and Aged Care, Australia
[log in to unmask]
Celine Noel <[log in to unmask]> on 21/04/99 12:52:34
Please respond to Celine Noel <[log in to unmask]>
To: [log in to unmask]
cc: (bcc: Prue Deacon/ITG/Health)
Subject: Re: Authority control/author affiliation
I am interested to see these two issues coming together in the DC
environment. Although I think there is value in authority control, it is
very expensive and not as effective as it might be even in the MARC
environment.
In the MARC-based authority file, we do spend a lot of time
differentiating authors with the same name. I think this is a worthwhile
activity but I don't think THE WAY we differentiate them (governed by
AACR) is very helpful: middle initials, fuller forms of name, birth date.
Are these the kind of things from which a user could recognize and choose
the correct entry from a long list of John Smiths? Not likely. On the
other hand, the author's institutional affiliation or field of work would
be very useful, even for the catalogers trying to be sure the John Smith
in front of them isn't already in the authority list (differentiated, but
not in a recognizable way) as:
Smith, John (John R.) [or any other middle initial for that matter]
Smith, John, 1936- [or any other birth date in the same century]
or even (God forbid!): Smith, J.
Author affiliation (or other work-related data about a person) won't work
for everything but if it's readily available and easy to record at the
time of metadata creation I think it is worth doing.
Celine Noel
UNC-Chapel Hill
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|