On 3/9/2011 3:25 PM, [email protected] wrote:
The alphabetical sequence (or possibly another systematic sequence,
but I can't think of another principle of organization) is essential
for maintenance of the catalogue,
Alphabetical sequences are no longer the only way our interfaces present
things to users.
One may wish to present all the records belonging to a certain work, in
a list by date. Perhaps the user wants only those in a certain language,
or perhaps not. Etc, there are other exclusions the user may or may not
want.
The Uniform title, while theoretically uniting records that belong to
the same work, does so in a way that can _only_ be used to unite records
by giving the user an alphabetical list of "uniform titles". Because it
is very carefully constructed to be optimized for that purpose, it is
very difficult or impossible for computers to reliably tell if two
distinct completely constructed uniform title strings actually represent
the same creative work, or not. There are many many different rules for
construction of a uniform title, which would need to be reversed (to go
from uniform title) and then encoded into software, with complicated
inter-relations. It is very hard and in some cases impossible to
automate this "reverse" process of getting semantics back _out_ of a
uniform title.
What is a 'uniform title' exactly, the user may ask? Well, it's just
this thing we assign so that records that we think should be grouped
together get grouped together when you put all the uniform titles in
alphabetical order. (Sometimes we think they should be grouped together
because they are all part of the same work, sometimes other reasons.
Such as the infamous Author, Some. Selections. )
So again, I applaud RDA for trying to disentangle the semantics of this.
The days are gone where the _only_ thing we creating cataloging data for
is to present alphabetical lists of headings.
to secure the necessary likeness and
differentiation among entities that are either the same or distinct.
Nothing confuses end-users of data as much as pointless distinction or
incorrect assimilation. I guess (in the light of the current
discussions of recording of series data) that it confuses computers
and programmers too.
Hal Cain
Melbourne, Australia
[email protected]
----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.
