First, thanks to everyone--especially Karen, Jörg, Amanda, and Amber--
for schooling me so thoroughly about current BIBFRAME thinking and
initiatives. My deliberately provocative initial email has elicited the
thoughtful responses I had hoped for.
While I can’t respond point-by-point to everyone, I wanted to follow-up
briefly, beginning at the “end,” so to speak, with the message sent by
Sharon this morning, in which she says:
“The whole point of RDA and BIBFRAME is moving to Linked Data community
for better information retrieval and connectivity on the Internet.”
This, I think, we can all agree on.
What troubles me (and this will be a major thrust of my presentation) is
the vast disconnect I see between two fundamentally different kinds of
discovery, or, if you prefer, use cases for the retrieval and use of
As someone who has worked in an academic research library as (mostly) a
cataloger for 27+ years, I realize that, with the dawning of the
Internet age, our collections have become, over time, simultaneously
more discoverable and less accessible. There has been an enormous shift
from print to electronic, and whereas anyone can walk in off the street,
discover, and access our print content (though not everyone can check it
out), our electronic content, for which we spend millions of dollars
every year (not to mention the costs of maintaining it) is accessible
ONLY to patrons affiliated with our university (with a few minor
exceptions). As our collections have moved online, we have unwittingly
contributed to the de-democratization of information.
Let’s imagine a time in the (near?) future in which BIBFRAME and Linked
Data principles have been fully embraced and implemented by the library
community (as well as others); systems have been built and tweaked to
accommodate the new data models and linkages.
Will the content discoverable in the new ecosystem be accessible to
users? Or will they be continually crashing into firewalls?
Finally, I would like to mention what I think is perhaps the most
extensive and thorough implementation of linked data in the
libraries/archives context: data.bnf.fr, created by the Bibliothèque
nationale de France. I mention it for a couple of reasons: 1) it
leverages existing UNIMARC data (FRBRized using software written
specifically for that purpose) and 2) the BnF--and French libraries in
general--have not adopted RDA.
Why have they not adopted RDA? Quoting http://www.transition-
“l’abandon pur et simple des normes actuelles de catalogage françaises
pour ce nouveau code de catalogage international s’avérerait coûteux,
contre-productif, voire constituerait une régression."
[purely and simply abandoning current French cataloging standards for
this new international cataloging code would be costly, counter-
productive, and would even constitute a regression.]
Further on, they note:
“Adopter RDA en l’état constituerait donc une régression par rapport à
ce qui fait la force des grands catalogues français, à savoir
précisément la gestion de ces liens."
[Adopting RDA as-is would therefore constitute a regression with respect
to the strong point of the large French catalogs, to wit—the management
of these links.]
The links in question are the ARKs connecting their bibliographic
records with their authority records. They have effectively been using a
form of linked data for 40 years.
[Two brief asides: for a skeptical view of RDA, see Karen Coyle and
Diane Hillman’s opinion in D-Lib Magazine from a decade ago (!):
ref=SaglikAlani.Com; also, Amber, to respond to your question about
failure of many libraries to adopt RDA—there was a recent straw poll on
AUTOCAT, with the results clearly showing that many smaller libraries
had neither the time, the resources, nor the will to move from AACR2 to
BUT, and this is a major point (now quoting Gildas Illien, formerly at
the BnF and now at the Muséum national d'Histoire naturelle):
“Data.bnf.fr does not mean to replace the existing catalogs and other
silos it exploits, but to provide some ‘glue’ between them.”
In other words, at least in the case of data.bnf.fr, neither the non-
RDA, UNIMARC-based catalog (http://catalogue.bnf.fr/index.do) nor the
BnF’s silo for digitized collections (http://gallica.bnf.fr/) goes away
as a result of the implementation of Linked Data.
This concerns me because cataloging/metadata departments, at least in US
academic libraries, are losing resources to other initiatives, even as
they are being asked to create and maintain metadata not only for
catalogs, discovery layers, digital repositories, and institutional
repositories, but now, also, for Linked Data projects.
Whence some of my skepticism of RDA, BIBFRAME, and Linked Data more
broadly. Too many generals, too few foot soldiers.