My understanding is that the next generation of OCLC's cataloging software will be able to control the terms used in 3XX fields of authorities and bibliographic records. I've certainly raised that as a need in discussions with folks from OCLC and have been told that the successor to Connexion will be able to do this. I'd like to see ILS vendors and authority control vendors also implement authority control over these fields. Hopefully someone from OCLC and various vendors is reading this and might be willing to comment on their plans?
University of Washington Libraries
How useful are the terms we record in RDA metadata fields in name authority records? That is, the 046 field, and the various 3XX fields.
Given the time it takes to record, does anyone think this data is being used for anything, or is likely to be useful in any way? Not in theory, but in actual fact?
It seems to me that the main functions of a name authority record for a person are these:
To identify a person uniquely to a user looking at the authority record.
To provide a means of collocating bibliographic records with the correct person in resource discovery.
To provide enough information for automated matching in ISNI.
The additional 046/3XX fields are, as I understand it, intended to be used in displays in a way that no system does, and to provide a machine-readable means of achieving some kind of linked data goal. However, the fields are optional in NACO and are not included consistently. There is no requirement to use controlled terms, or to establish the terms in controlled vocabularies (SACO will not even accept proposals for new LCSH used in NARs). Terms that *are* taken from controlled vocabularies (LC/NAF, LCSH, LCDGT, etc.) are not maintained. So Iím uncertain of their value as potentially linked data.
Meanwhile, PCC is working towards an identifier-based model of identity management. ISNI matches data primarily on form of name, dates, and associated titles, and doesnít make use of the RDA metadata fields in NARs (affiliations in particular have performed poorly as factors in algorithmic matching of identities).
The first two goals above can also be achieved just by recording preferred and variant names, and the source information we have recorded in 670 fields since the year dot (and still record to justify the content of 046/3XX).
Itís often the case that by the time everyone has implemented a new thing, the paradigm has shifted again and much of the new thing is no longer useful.
Resourcing constraints suggest that we should look at these things quite closely.
(My opinions, not necessarily those of anyone else)
Authority Control Team Manager
The British Library
Tel.: +44 (0)1937 546104