Dear colleagues,
a few thoughts from the German speaking part of the
MARC Community about Proposal No. 2010-06, Encoding
the ISNI in MARC 21.
In general, we support the approach of representing
an ISNI in a subfield $0 and thus extending subfield
$0. It is a logical step to allow a $0 to carry not
only a r e c o r d number, but an identifier as
a whole.
There are some implications: Up to now we can rely
on the existence of a record "at the other end" of
a $0. There is a MARC record (or ist equivalent)
with a field 003 containing the part of $0 in brackets,
and with a field 001 containing the rest of $0.
And there may be MARC records describing the same
entity, containing a field 035 "System Control Number"
with a subfield $a with exactly the same content as
the $0.
With the new logic, either there is no MARC record
at all with 001/003 or 035 $a; or not yet; or there
is a record with e.g. a field 024 7# and a $a
containing the non-bracketed part of $0, and $2
containing the part in brackets.
We are satisfied to see that the brackets are still
there, giving a kind of attribute to the identifier
itself, either a code for the institution, or now a
code for the identifier scheme.
Fortunately, the list of "Standard Identifier Source
Codes" is not a very long one.
For reasons of consistency, we could imagine to take
subfield $w into account when discussiing $0. As
opposed to $0 (the "Authority record control number")
$w is the "Bibliographic record control number"
leading to a bibliographic record. We can imagine to
extend the extension of subfield $0 to subfield $w.
Some values in the list of "Standard Identifier Source
Codes" can be seen as clearly identifying an entity
that is described in an authority record, "isni" e.g,
while other values are related to a Bibliographic
entity, "ean", "isbn", "upc". So it may be logical to
allow something like "$w(isbn)1234567890123". On the
other hand, there are some subfields that already
transport specific identifiers.
And: there is this value "uri" in the list. So one
implication is the question: Could something like
"$0(uri)http://d-nb.info/gnd/118634313"
be possible? And how would that relate to the former
approaches to encode a URI in MARC?
Regards
Reinhold
--
Reinhold Heuvelmann
German National Library
Information Technology / Data Formats
Adickesallee 1
D-60322 Frankfurt am Main
Telephone: +49-69-1525-1709
Telefax: +49-69-1525-1799
mailto:[log in to unmask]
http://www.d-nb.de
|