With apologies for cross-posting.
After the recent discussion on the SACO list concerning LCSH proposals for new terms needed in name authority records, and PSD’s response, I’d like to develop this case.
When I reviewed the use of controlled vocabulary in LC/NAF earlier this year, 49% of 368, 372 and 374 fields had a source given in $2, and of those, 95% were LCSH. 100% of controlled vocabulary used by the BL was LCSH, and 94% of controlled vocabulary used by LC was LCSH (LC and the BL between us are responsible for around 40% of new LC/NAF records).
LC have provided some guidance in DCM:Z1 on the use of LCSH in NARs. Since we began creating RDA NARs at the BL in 2011, we have had many discussions with our own cataloguers, which resulted in further guidance to be found in the BL Guide to RDA Name Authority Records (RDA Toolkit, Global Workflows).
This illustrates that LCSH has emerged strongly as the controlled vocabulary of choice, in LC/NAF name authority records. That is understandable, as our cataloguers are familiar with it and need no further training, and desirable as it promotes consistency within LC/NAF, and with controlled vocabulary in bibliographic records for linked resources. It’s efficient for us, as our local copy of the LCSH file is linked to our bibliographic database, and can quickly be consulted.
We have found LCSH suitable to the task. The only significant issue has been not with its structure but with gaps in its vocabulary. “Kingdoms” was the first proposal intended to fill such a gap, but it was designed to be compatible with bibliographic LCSH usage, and literary warrant was adduced as well. We have not proposed any change to the structure or rules of LCSH.
Interestingly, the authority used as a pattern “Republics” has no usage or literary warrant given. LCSH is not a pure thesaurus, but records terms used by the Library of Congress in the past, and terms proposed by SACO institutions as needed. LCSH rules and patterns were not designed from scratch; the SHM is the result of an exercise in bibliographic archaeology conducted by Lois Mai Chan and others.
As a result, LCSH is inconsistent in structure across subject areas, and often contains terms that are less than ideal. However, it is the system that it is most efficient for us to use. The same objections raised to its suitability for authority records can be (and have been) made in respect of bibliographic records.
Suggestions have been made for other thesauri that could be used, or that might be developed. When I began my career at the British Library in 1991, the Authority Control team was putting the finishing touches to a subject system called COMPASS (Computer-aided subject system), that used terms mainly from Precis, arranged in a fresh thesaurus from which faceted strings were created. Theoretically, it was much clearer and more consistent than LCSH, but we dropped it because its potential users found it more effective to use LCSH, as the prevailing standard.
22 years later I find myself in the odd position of commending LCSH to its creators. I hope that PSD will consider accepting proposals for new LCSH needed in name authority records, as long as the proposals are consistent with LCSH principles for bibliographic usage. In principle anything that has a name authority record belongs to a class of things, or persons, that can also be written about.
Authority Control Team Manager
The British Library
Tel.: +44 (0)1937 546806