That's really sad, because there was a time when the real
>innovations in librarianship came from practicing catalogers
>(or at least people working on the principles of organizing
>information, such as Dewey, who is hard to pigeonhole in one
>specialty).
I agree with you and truly hope that catalogers will be able to
contribute their skills and knowledge to the development and
implementation of next generation catalogs. I don't expect most of us
to be involved in the same tasks we are presently doing 10 years hence,
at least in local library settings. Some will - I hope that scholar
level catalogers doing original work primarily for researchers will
continue - but many of us will be likely be working very differently.
An information system I developed in 1999-2000 demonstrates the
viability of Dublin Core, user tagging, and metadata harvesting blended
with LCSH and LC name authorities. Though I have not been involved with
it since mid-2001, the Texas Records and Information Locator (TRAIL)
(http://www.tsl.state.tx.us/trail/) continues to provide search and
retrieval of tens of thousands of documents from more than 180 Texas
state agencies. My background as a cataloger provided a basis for
developing what was, at that time, a fairly innovative approach
including:
- use of tagging by agency staff as a primary means of keyword topical
access (3-5 tags or tag phrases per document) as well as summaries for
all documents
- use of a subset of LCSH as a framework to provide common "collections"
of like documents (1-2 LCSH terms per document)
- use of Dublin Core as the basis for the cataloging records. In our
first couple of years, TRAIL was the largest contributor of Dublin Core
records to OCLC's CORC project
- fully functional Z39.50 mapping of DC to MARC providing the capability
to search across TRAIL, the Texas State Library bib catalog, and any
other Z39.50 capable database (RDF, XML, and other means of mapping
across databases were just emerging so I used what was available)
- linking from any document record to the online document itself, as
well as to detailed information about the agency
- a commitment (at that time) to continually develop and improve the
capabilities of the system and the interface instead of remaining static
(continually is the key word in that phrase, BTW)
- utilization of ongoing focus groups, comments, and session log
analysis to continually improve the system
As one individual administering the system, I trained and coordinated
the work of 200+ agency staff who created Dublin Core records (now
imbedded in documents when possible). I actively mined the search logs
to add keyword tags to relevant records, made corrections, and planned
and implemented system enhancements.
Looking at it now, it is soooo last century (no offense intended to the
present administrator) but after 8 years of heavy usage (it was the most
used or close to the most used web resource in Texas State government
during the time I was there), it continues to serve as an example
(albeit out-of-date) that alternate approaches based on cataloging
principles and practices can effectively serve users.
Given staff and resources, I could easily have chosen a traditional
cataloging approach. If I had done this, I believe the system would
have taken many years to get off the ground (I planned and implemented
it in less than 2 including a very robust collection across the majority
of available web resources in Texas state government) and would not have
served users any better, if even nearly as well, as the approach I took.
On many, many occasions I experienced instances where use of tagging
provided links to documents that satisfied user needs when LCSH would
not have done so. Use of LCSH did provide for common collections of
documents and topical interface with traditional catalogs using common
terminology. If maintained currently, LCSH would also be providing
faceting as well.
So when I speak (as I do often in this forum), it is from the
perspective of a present-day cataloger (about 20 years of cataloging
related experience) currently doing a decent job of traditional
cataloging in a public library setting (I love public libraries and am
an avid reader) but also from the perspective of someone who has
successfully planned, developed, implemented and administered a fully
functional bibliographic information system (next generation in the last
century) that fully utilized tagging before the term had been developed
and utilized some other themes and ideas that we sometimes discuss here.
And the world didn't fall apart in Texas just because electronic
documents were "cataloged" in a different manner than a traditional
approach. TRAIL is not perfect, then or now, but it remains viable and
useful for users.
And I've been on the waiting list at our library for David Weinberger's
book since I first encountered it.
Allen Mullen
Received on Sat May 19 2007 - 08:56:18 EDT