While I appreciate the concerns over undifferentiated records, I'm a bit
alarmed at the idea that they might be banned. (Not being at present an
OCLC user, I won't comment on linking or not linking in Connexion.)
Names that obviously or very probably represent different people, but
for which there's insufficient information to construct distinct forms
of name, are a fact of life. Users (both librarians and others) are
going to continue to search for them, and some of those names will
require references; it's also beneficial to record information found and
not found; so it seems plain to me that authority records are still
called for. Maybe we can be smarter about the way we use the
information we have.
I agree with Mary Charles that LCCNs should change when the status of a
name record switches from differentiated to undifferentiated; we expect
a one-to-one relationship between ID number and the entity designated in
the record. This is a situation quite different from when the
established form of a name is changed: the referent of the record, the
identity it designates, is altered.
I would be happy to see the use of designations in $c extended to break
at least some of these undifferentaited names. But I have two
reservations, or at least points for consideration. First, the more
"permanent" in nature the designation is, the better -- I wouldn't like
terms such as (to create a fictitious example) "Ph.D. candidate,
University of Melbourne"; whereas titles such as "Rev." or "Archbishop
of Melborune" are more likely to be permanent (though not necessarily:
the previous Catholic archbishop of Melbourne is now archbishop of
Sydney and a cardinal -- not that he's a case of a name conflict).
Similarly we should be cautious about designations which are related to
only a part, especially a small part, of a person's life and activity.
Second, in most catalogue index displays, qualifiers added to a name
affect its position in an alphabetical sequence, thus may make them
harder to identify and retrieve -- a situation which generates its own
problems and gives rise to errors.
We already have, courtesy of OCLC, a notification mechanism for names
modified by addition of death dates. If we had also a notification
service for other one-to-one changes (same identity, altered established
form, and for changes between differentiated and undifferentiated
status, we would have better tools for keeping abreast.
It could also be argued that undifferentiated names should be designated
as "provisional" because the work can't be completed. What practical
difference that would make I don't know.
One other troublesome area (for which I often send reports to CPSO) is
where there are multiple persons with the same name, and one (though
unique and technically differentiated) carries no further qualification.
One I addressed recently to CPSO, for removing arroneously attributed
works from the unqualified name, was: Dods, Marcus (n 84054382) -- 670
shows he lived 1918-1984. There are others (all established) of the
same name: Dods, Marcus, 1786-1838; Dods, Marcus, 1834-1909; Dods,
Marcus, 1874-1935. In this case, allowing the unqualified form to stand
beside forms with qualifiers makes more work for the catalogue user, and
generates errors by cataloguers too (bad, but it happens repeatedly, in
LC records and others).
One more thought: I find problems with names established on the basis of
a single occurrence -- especially when that occurrence is in an
unpublished work, such as a thesis. It's my view that names based on
unpublished sources should always be provisional; and that names based
on a single occurrence (especially CIP information) should be
provisional when established, and remain so for, say, 10 years. At
least we would be alerted, when we had to investigate, that the
information needs to be reviewed.
Joint Theological Library
Parkville, Victoria, Australia
[log in to unmask]