McGrath, Kelley C. wrote:
Karen,
... I find the idea of a recordless view intriguing and presumably much more
flexible.
Karen Coyle had said:
What worries me most about the FRBR WEMI view in which each entity is
a record is that it places a nearly impossible burden on the
cataloger. Which is why I'm exploring the possibility of a
recordless view -- which would consist of short statements (Jane is
author of Book) that are each valid, and can be combined with other
statements to build up to a complete bibliographic description. I
don't know yet if this is possible. It would use semantic web
concepts, not the RDA scenario 1, but perhaps scenario 0. It assumes
an open bibliographic environment where statements can exist with
metadata contributed by others.
About any particular book, there can be many statements out in the
open world of the Web. Provided there is a stable, reliable, unique,
universally used identifier, going with every suchj statement, you're
very nearly there. The ISBN and ISSN are not quite that good, but the
best we have, and they do already play the part of that identifier
in many practical scenarios.
In the paper world. we also had (and have) many statements about many
documents, scattered throughout the literature - those, however, could
not be found or collocated at the push of a button. It was an
intellectual endeavor, and a very time consuming one, to collect
statements about a book.
The classical OPAC, of course, gave you just the record and nothing but
the record that was in your database. Contemporary OPACs do already
give you a lot more, they may link to any number of resources outside
the catalog, and they may or may not pull in data about the book from
other sources, like ToC or cover image. Following the links, you may get
pointed to even more sources saying more things about the book in
question. The links, notably, need not be part of the book's record!
Provided there is the identifier - presently the ISBN/ISSN - links to
all sorts of services and sources can be constructed by the OPAC
software on the fly, and changed any time.
Identifiers, and we've known that for a long time, are therefore
key for navigation, although search statements for other databases
can be constructed out of author/title data, but...
Identifiers, of course, serve the known item search only. We also have
to consider the collocation search (for other expressions and
manifestations) and the subject search. But will every library need
to have pertinent statements in their databases when such statements
are available and accessible elsewhere?
So, given the infrastructure that is in place now, and which is
growing and improving without our moving a finger, what will be the
minimum statement(s) a library needs to make about a document in its
collection? And what functions will the library have to provide
to establish a service that achieves more by doing less? And how
much less (or more? or different things?) will that be than what RDA
calls for? And where, coming to think of it, is the need for a complete
bibliographic description as we know it? (And what would complete
mean, anyway.)
B.Eversberg
