Kevin said:

>Rather, a box in which the name could be input, and another box next
>to it that would have the relationship.  A likely scenario would have
>the possible relationships appear in a drop-down menu.

If Bibframe makes the personal/corporate/conference/uniform title main
entry (aka preferred entry) distinction, there would have to be four
boxes, or a second dropdown window to select the nature of the name.

I could see that working, particularly if you had the option of
relationship terms (MARC $e) or relationship codes (MARC $4) for the
dropdown window, and if that added the term or code to the heading.  

I don't see how it could determine indicators, for example in
AACR2/MARC whether surname or given name, or whether a corporate name
is direct or subordinate; perhaps Bibframe won't make those
distinctions?  (In that case a MARC to Bibframe crosswalk would be
possible, but not the reverse.)

I'm no fan of relationship terms or codes, but if they obviate those
display labels, I could warm to them.  Those labels are either wrong
(e.g., criminal defendant called "Author") or too general to be of any
use (e.g. "Person").  They take up display space needed for real data,
particularly with more wordy RDA records.

There is still the problem of legacy records lacking the relationship
terms or codes.

   __       __   J. McRee (Mac) Elrod ([log in to unmask])
  {__  |   /     Special Libraries Cataloguing   HTTP://
  ___} |__ \__________________________________________________________