> And while I would argue against adding death dates to open birth dates as
> an incomplete/unpractical maintenance solution, changing "DATE-" to "b.
> DATE" could be done automatically by a machine flip to retrospectively
> "correct" headings already created with open birth dates.

I would like to respond to a very small portion of Diana's message.
A machine flip in one machine may be a simple process. When you
consider how many unlinked machines may have a particular name
heading, you have to remember that they would all have to do this
simple machine flip and that many of these libraries would not send
a version of their flipped records to the utilities where it would
be copied again and again. I know not all headings have many
occurrences but many do. Our construction of headings should be
conservative as should our desire for global change.

The last time MARBI discussed the "why not use b. for folks
presumably dead?" issue, the problems with interfiling open dates and
b. DATE entries were considered significant enough to "defeat" the
proposals.

Sherman Clarke
NYU
[log in to unmask]