Kevin said:

>Transcribed data and controlled data serve different purposes, and
>they cannot be substitutes for each other.  Neither is more important
>than the other.  Yes we need to have linked data, but not at the cost
>of losing transcribed data.

I strongly support providing lacking data in transcription.  RDA now
allows jurisdiction in imprint, but not "[sic]" or clarifying words
such as "[proceedings]" in titles, or functions in statements of
responsibility.  Because of legacy records, Bibframe should have slots
for those inclusions however.

Before the elimination of 440, that was a field which served both
purposes Kevin gives.  When the form on the item and the standardized
form of any data is identical, having both does seem a bit redundant.  
MARBI obviously agreed with Kevin when they divided the two functions
between 490 and 830.

Titles are both transcribed and access points, but not (in the case of
245) controlled.  I assume transcribed data in Bibframe will not be
controlled, and that there will be no double usage as formerly
represented by series (440), but that transcribed data can be an
access point as represented by tattle proper.

As mentioned before, if the standardized form is to be automatically
derived from the transcribed form, the transcribed form may need
enhancement, e.g., supply jurisdiction if not on the resource.

MARC has provisions which go beyond AACR2 or RDA (e.g. fixed fields),
and so might Bibframe (e.g. supplying jurisdiction).  What is encoded
is not solely a matter of the rules.

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