On 2/3/2011 1:58 PM, Karen Coyle wrote:
Where does "controlled access point" fit in? Are there defined access
points in RDA that are not authority controlled? (I guess titles
proper aren't).
The title proper is not a controlled access point; I'm not sure that it
is defined as an access point at all. I think that the concept of
"authorized access point" is limited to things that (in today's record
structures) can be encoded both in the bib record and in an authority
record -- it is the latter that authorizes the former, so to speak.
RDA does not use the term "controlled access point". If I were going to
use that term, I'm not sure whether I would equate it with "authorized
access point" or with "controlled vocabulary". If the latter, then
there are lots of terms in controlled vocabularies for which we do not
(yet?) have authority records.
We seem to have these categories in the rules:
- descriptive information (mainly text)
- authoritative/controlled access points
- other access points (titles)
- ?? Linking information?
Mostly, these are not categories that have any reference in RDA. These
are categories that are based either on earlier rules, on MARC encoding,
or both.
In RDA terms, there are two categories: attributes/elements/properties
and relationships. In RDA, we have descriptive information (and yes it
is mostly textual) that are recorded in elements that identify each of
the FRBR entities. We also have access points (both authorized and
variant) that are constructed by combining relevant elements into a
unique text string; the authorized access point is one of the ways in
which relationships can be recorded (an identifier is another way).
The linking information is a type of relationship. There is nothing
special about the linking relationships in RDA, but in MARC there is
distinctive encoding in the linking-entry fields.
Oddly enough, I don't think that there are specific instructions about
constructing access points for titles as such, only for constructing
access points for works, expressions, etc. (which may in fact turn out
to be titles if the name of the creator is not included in the access
point for the work, etc.). I believe that RDA recognizes (without
stating it explicitly) that an application may provide access based on
ANY element; but RDA only includes instructions on constructing access
points for the FRBR entities.
MARC, to my mind, is a confusing element in the analysis of RDA data.
It was designed before RDA or the models it is based on were created.
RDA (and FRBR, etc.) takes a very different approach to data
categorization than does MARC. Personally, I think the RDA/FRBR
categories are a step forward towards rational data, but that is only an
opinion.
MARC adds the fixed field data, all of which are data of some sort,
and the linking fields.
Regarding the fixed fields, many MARC fixed-length data elements are
coded equivalents of RDA elements and may be used as alternatives.
Beyond that -- as Tom Delsey documented in his analysis of MARC -- there
are a variety of administrative tasks, entities, and elements that MARC
has to support that fall completely outside the scope of RDA.
John Attig
Authority Control Librarian
Penn State University
[email protected]
