I had not before contemplated round trip mapping of data from MARC21 to BIBFRAME. But I rather like the prospect that data with the potential for coding in multiple places in MARC21 could be roundtripped through
BIBFRAME and reliably mapped to all appropriate places in MARC21 (even though I hope to be making the transition in one direction only). It is one of the murkier aspects of our MARC data. It would be nice if clean coding in BIBFRAME, using whatever control
mechanisms are implemented, afforded clean transfer back to MARC21 in both coded and textual forms. This would introduce (greater) redundancy in the MARC21 records, but the redundancy would be a) uniform and b) in a form that were better actionable. And
it would be just as nice, if not preferable, that the multiple presentation options in MARC21 were regularized. (I am thinking of things like the need to code illustrations in the MARC 008 and record illustrations in the 300 – if we had a mechanism to encode
illustrations using a vocabulary registry, then the instance of either in MARC21 could be mapped to the registry in BIBFRAME and then mapped back to both in MARC 21.)
John F. Myers, Catalog Librarian
Schaffer Library, Union College
Schenectady NY 12308
518-388-6623
Sally McCallum wrote:
Based on our experimentation thus far at LC, I would not expect the trips either way to be lossless, however, in some cases, a trip could actually add clarity to the source data.