I have a few thoughts on both Jonathan's comments and on Diane's.
Jonathan Rochkind wrote:
> But, there is indeed a 'tragedy of the commons' issue there. Right now,
> many people are already making those record improvements---because they
> believe they have to, if they want to get those improvements. If such a
> technical infrastructure allowing better sharing existed, would all
> those institutions stop making the improvements, hoping that 'somebody
> else' would, and they could just take advantage of them for cheap? If
> everyone is waiting for 'somebody else', then nobody's doing anything,
> indeed.
This is some of cataloging's unspoken dirty laundry. Already happening.
We ALL sit on materials in hopes that somebody else will catalog it
first. It's not a problem so long as we're not all sitting on the
*same* materials. Each institution has its own priorities and does
original cataloging on those priority materials and we hope that our
priorities are disparate enough that we can all take advantage at the
mass network level.
Jonathan Rochkind wrote:
> I thought you said you were _already_ making them, for the presumed use
> of your own users? If you are already doing that anyway, would you
> really mind sharing it with others with no additional gain except the
> benefit of the library community (and the fact that you will mutually
> benefit from this strengthened community, in a sort of generalized
> reciprocity).
I am happy to share data and I am happy to do so freely. But, at the
University of Virginia (as just one example), we load records in from
pretty much any source that we can acquire them. That means, in our
workflow, the records are already in my local catalog when we touch them
for the first time. Under the current OCLC Enhance Sharing paradigm,
upgrading records in Enhance is double-the-work. If we could fix the
record locally and push-button send it to OCLC, we'd shard our upgrades
happily. I want our edits to benefit the larger community, I'm all
about the greater good, but I'm less willing (a lot less willing) to
slow down our workflow to do double-the-work in OCLC. To me, that is an
explicit cost. It's no longer about freely sharing our edits, it's now
about adding cost into the process. If we continue to maintain local
stores of records ('tis the question), we need a whole new model for
cooperative enhancement.
Diane Hillman wrote:
> It is certainly very expensive if the only way you think about
> "touch" is our old one-at-a-time way of doing things. However, there
> are other ways to approach the problem. There are automated methods
> to add or change or "enhance" information that we have not taken much
> advantage of in the past. Some of these methods have been used by
> OCLC and (formerly) RLG as they worked with the data, others are
> being developed outside of the 'mainstream' distribution channels of
> traditional libraries.
There are a few assumptions built into your logic here that, in my
opinion, need light. There was a bit of discussion at the FoBC the
other day about these discussions not being about "reducing" costs, but
rather about "shifting" costs. Get the publishers more involved at the
beginning of the supply chain (shift cost to the publishers), get OCLC
to pick up more slack, get authors (photographers, in the National
Geographic Society example) to submit metadata, etc. While I actually
very strongly agree that we need to push descriptive cataloging to the
beginning of the supply chain and that we should enhance records over
time, and while we at UVA have been a long proponent of "the perfect is
the enemy of the good" in creating cataloging records, the kinds of
tools that are needed to enhance records in an automated way over time
COST MONEY. Either in software costs or, open-source, in development
costs. And since most of our ILS systems don't provide adequate support
in the environment we currently live in, for catalogers down in the
trenches it is really hard to envision a time when our vendors will
provide adequate support for mass data manipulation. Not that we don't
want it, it's just hard to really see it happen in the economic
marketplace in which we currently find ourselves. OCLC (as usual) is an
interesting example here. They obviously have high-end sophisticated
tools for data manipulation. They've never shared them with the rest of
us. They have highly trained staff that do sophisticated processing of
large quantities of data. Yes, we can rethink the Enhance model, but is
it feasible that they would consider mass data manipulation to also be a
"cooperative" task? If we had good tools, it would be much easier to
imagine a time when we could dump in mass quantities of more
minimal/inadequate/provisional (we do this all the time for acquisitions
order records) and enhance/enrich/correct them later programmatically or
manually as necessary. But, even in my smaller context, I want my
cataloging staff to be able to enhance/enrich/correct data in mass (just
on a UVA scale) and getting resources put towards tools that would
facilitate that is very very difficult. It's not cheaper, it's just a
way of shifting cost. I agree that this is where we need to go, but I
think we do need to say out loud that this *also* will cost a chunk of
change.
Message too long, thanks for reading!
Erin Stalberg
Head, Cataloging Services
Chair, Metadata Steering Group
University of Virginia Library
stalberg_at_virginia.edu
(434) 982.2854
Received on Thu Jul 12 2007 - 07:48:09 EDT