I have followed this discussion with much interest. I agree with Adam Schiff's comments below.
I can appreciate that it is worthwhile, whenever possible and desirable, to replace a personal name heading with a birth date with a heading that includes the death date. This is extra work and doesn't really improve authority control but as a public relations exercise, it has merit.
However, I wish to state in the strongest terms I can, that to change the RI which asks us to include the birthdate (or dates) when known would greatly increase our work (here in Singapore) and would most likely result in degrading what authority control we have in the area of personal name headings.
As many of you will know, Singapore's population is largely ethnically Chinese. The Chinese in Singapore draw their surnames from a pool of the '100 common surnames' and now often combine these common surnames with 'English' given names. We go to some trouble to find out birthdates of these authors just to save precious resources later when another author by the same name is published. In addiiton, many of the 'Chinese' given names (dialect, not Pinyin) are used extensively. I would love a dollar for every name heading that we've established that has 'Eng' (an evergreen popular Hokkien given name that means 'eternal') in it.
We already have difficulties with the headings used in LC bibliographic records that don't reflect a date and where we know that sooner or later a conflict will emerge. At this point we are content to put the additional date information in a 670 and replace later when the conflict does emerge but a limited diet of this type of activity is definitely in order.
The study Gary cited in which it was found most of the headings that contain dates don't involve conflicts in future begs several questions, such as length of time before conflicts arise; and, the resources invested in breaking conflicts with the review of large bibliographic record files.
SILAS (Singapore Integrated Library Automation Services)
Adam Schiff wrote:
I agree with Gary's proposal except for the last paragraph of it.
Instead, I would rather see the birth date given in a newly establishe
heading any time it is known. This would obviate the need to have to
change that heading later when it comes into conflict with another
heading. Otherwise, I think Gary's proposal limiting the addition of
dates on already established headings to these situations is a good one:
1) any kind of date or dates need to be added to previously established
heading to break a conflict with another established or newly established
2) death date added to established heading with an open date of birth (so
we don't look foolish; those libraries who receive notifications or new
copies of changed headings should not have much extra work to do for a
change like this if the change of the authority heading automatically
propagates to the corresponding bib. records)
3) open data of birth on established heading changed to "b." date
if known that the person has died but date of birth unknown (again,
changed so that we don't look foolish and those libraries with an
authority service that receive notifications or copies of changed records
that already matched one of their bibs. should have a fairly simple time
getting their matching bib. records changed)
Adam L. Schiff
University of Washington Libraries
Seattle, WA 98195-2900
(206) 685-8782 fax
[log in to unmask]
StarBiz - Visit http://www.starbiz.net.sg