Because of the higher level of granularity in MARC (over EAD), it might
be more appropriatle - if you are making new records - to start with
MARC records and then simplify them to EAD. Just in subject fields
alone, MARC allows for several distinct subfields. EAD does not. So
additional work - post transformation from EAD to MARC to further
subfield will be required.

And Michael is also correct, encodinganalogs are extraordinarily useful
in such mappings precisely because EAD's elements are more generic than

Liz Shaw

Fox, Michael wrote:
> I would not characterize encodinganalog as a relic since it is intended
> to be used in just this sort of process.   Since the structure of MARC
> does not parallel that of EAD in all respects and because MARC is
> sometimes more granular the EAD, an "internal cross-walk" may be
> required.  How else, for example, can one predict whether <persname> is
> mapped to MARC 100, 600, or 700?   Either the encodinganalog or role
> attributes could be used to accomplish this specification.
> Michael Fox
>     -----Original Message-----
>     *From:* Encoded Archival Description List [mailto:[log in to unmask]]*On
>     Behalf Of *Yann Nicolas
>     *Sent:* Thursday, July 15, 2004 8:22 AM
>     *To:* [log in to unmask]
>     *Subject:* Spam: EAD => MARC
>     Bonjour,
>     we plan on creating MARC records from EAD instances.
>     We want our conversion method to fulfil these three requirements :
>     1. our two steps process will first convert from EAD into some
>     XML-ized MARC and then from this XML-MARC to ISO 2709 ;
>     2. the first conversion will use an XSLT stylesheet ;
>     3. this XSLT transformation will be able to select any EAD element
>     and attribute, not only elements qualified by an 'encodinganalog'
>     attribute.
>     Therefore, I have two questions : the first is motivated by
>     curiosity and the second by... interest.
>     1. What is the /raison d'etre /of the 'encodinganalog' attribute ?
>     What is the advantage of an internal crosswalk over an external one
>     ? Is it just a 'relic' of old times, before the XSLT and XPATH
>     possibilities ?
>     2. Is there somewhere an available XSLT stylesheet to convert from
>     EAD into, by instance, MARC 21 XML schema (Library of Congress) ?
>     It would be a useful beginning point for us.
>     Thank you
>     Merci
>     *Yann NICOLAS
>     ABES - Agence Bibliographique de l'Enseignement Superieur       (France)
>     Metadata librarian
>     */Service /: Produits et Services
>     /Dossiers /: Portail Sudoc, Theses electroniques, CGM
>     04 67 54 84 25
>     mailto:[log in to unmask]