NSF/JISC Repositories Workshop
Ken Hamma,
J. Paul Getty Trust
March 28, 2007
Download: PDF Version WORD
version
We recently went through an exercise to see how difficult
it would be for Getty collections to expose non-bib cataloging
information (metadata) in the OAI Protocol for Metadata Harvesting
and digital surrogates (resources) in a standard File Transfer
Protocol for an aggregator to harvest. The OAI harvesting
worked in the usual way and a simple URL reference in each
XML record triggered the subsequent FTP download and association
of resources with metadata. After a few bumps this worked
well and, we are convinced, is a simple low-cost alternative
(until ORE) to the cumbersome and resource intensive models
for contributing to aggregate resources that were part of earlier
efforts like AMICO, the Art Museum Image Consortium. (In
the end, a single export routine will makes information and
images available to OAI exposure and to publication on getty.edu,
and makes indexing available to search engines.)
If one wants eventually to have an environment in which it
is easy to find, for example, all paintings by Rembrandt Peale
and associated archival materials, the strategy of making that
possible has to envision the participation of many small and
under-funded institutions. Because of that we assumed
it is critical to place the bar for cataloging and technical
investment as low as possible without impairing simple online
discovery. We also began to explore models where key
aggregators like OCLC would host open OAI servers as well as
publicly available services like vocabulary assisted searching,
where application vendors would provide an export mapped to
the XML schema, and where big institutions would provide services
to smaller ones in regional or intellectually affiliated consortia.
This provides a model for the useful exposure of potentially
all repositories of unique works in museums, archives, libraries,
which are among the key data repositories for scholars in the
humanities. Moreover it does so at the network level
and in a way that is predictably available to any institution
or individual interested in exploring and aggregating for any
purpose, from large scale service providers (ARTstor) and portals
(EUBAM), to very large scale resource aggregators (WorldCat),
to very small and individual research projects or single classroom
teaching.
Why won’t this work today.
- What OAI refers to as resources most archives and museums
refer to as assets, and assets are expected to make money. This
has a long tradition in museums, which require even scholarly
publications to pay significant image licensing fees, and
perhaps not as long a tradition in archives from which the
success of a PBS documentary inevitably elicits groans of
wanting to share in the money. This is also completely
understandable in institutions that, aside from a very few,
are habitually under-funded and under-staffed. It is,
indeed, an explicit government expectation in many countries
that national museums “use” their collections
as assets in this way. All of this leads to severe
gate keeping at odds with anything that has a scent of openness.
- Even
with (sometimes especially with) the research community there
is a mistrust of intentions and a feeling among museum professionals
(archivists are getting over this but still struggle) that
they must guard over the proper use of resources (assets). This
covers a wide range of behaviors from knowing who has got
a resource to knowing how it is being used, e.g. not being
published with inappropriate cropping or insufficient color
management. It is clear to everyone
that no such controls truly existed in the analog world,
but the ease of high quality digital distribution inevitably
leads directly to another reason for gate keeping.
- For authors
a different sense of ownership attaches to even a modest
set of cataloging information. This sensibility
is strongly grounded in the traditional print model of collection
catalogs in which the authors’ names and reputations
(and in academic settings tenured career opportunities) are
inextricably bundled with the entire scope of information,
no matter how trivial, no matter how rich. Many people
recognize, once pushed, that ensuring attribution and authenticity
in a digital environment is sufficient, but the environment
itself is not sufficiently well built out that there is much
trust. Exceptions exist.
- Institutions acting in the
same area exhibit an extraordinary lack of imagination. Deliberate
limitations on quality and quantity in available data are
seen to ensure the viability of traditional modes of scholarly
communication. (This
can be either a principled decision or one on funding.) And
these traditional modes seem very satisfying especially when
the network goes down. Never mind that the network
has not failed for decades; we fondly remember the days when
it predictably did on a weekly basis and that reinforces
concerns expressed about the future of those bits and bytes.
A
statement released by the Association of American University
Presses in February of this year contains the blinkered observation: “The
increasing enthusiasm for open access as a model for scholarly
communication, which grew out of the pressure to relieve the
financial burden on libraries of maintaining subscriptions
to STM journals, presents new challenges and new opportunities
for university presses. In its pure form, open access calls
for an entirely new funding model, in which the costs of publishing
research articles in journal are paid for by authors or by
a funding agency, and readers can have access to these publications
for free.” What body other than the university
or research center itself a) employs the researchers, b) provides
subventions for press runs of fewer than 500, and c) comes
up with the budget for the library to buy the subsidized but
still expensive books for the researchers they employ? If
funding is not the real issue – there is a lot of money
in that system – could it be that we are expecting too
much in the way of critical path decisions on institutional
policy and intellectual property from publishers comfortable
with or desperate to preserve a traditional model, and librarians
just sufficiently well funded to be comfortable with their
own and publishers’/distributors’ traditional roles?
- And
because we are not sure we want to, for all the reasons above,
it is difficult even in well-funded institutions to find
the money for item level cataloging and digitization of large
bodies of resources. Smaller institutions
don’t even get this far, stopped in their tracks by
the cost of owning technology and managing commercial applications.
Flavors of these issues can be found nearly everywhere one
looks, and I don’t mean to be pointing a finger specifically
at humanists, let alone those I work with. This is just
one arena where the characteristic of openness – a characteristic
that efficient networks and interoperability assume – makes
opportunities seem scary.
I began thinking about this short text not with our OAI exercise
but with a position statement from the Wellcome Trust. This
won’t be new to a JISC/NSF audience, but I would hazard
that it and similar position statements have not yet appeared
on the horizon at, say, NEH or NEA.
The mission of the Wellcome Trust is to foster and promote
research with the aim of improving human and animal health.
The main output of this research is new ideas and knowledge,
which the Trust expects its researchers to publish in quality,
peer-reviewed journals.
The Wellcome Trust has a fundamental interest in ensuring
that the availability and accessibility of this material is
not adversely affected by the copyright, marketing and distribution
strategies used by publishers (whether commercial, not-for-profit
or academic).
With recent advances in internet publishing, the Wellcome
Trust seeks to encourage initiatives that broaden the range
of opportunities for quality research to be widely disseminated
and freely accessed.
The Wellcome Trust therefore supports unrestricted access
to the published output of research as a fundamental part of
its charitable mission and a public benefit to be encouraged
wherever possible.
Specifically, the Wellcome Trust:
* expects authors of research papers to
maximise the opportunities to make their results available
for free and, where possible, to retain their copyright
* will provide grantholders with additional funding to cover
open access charges levied by publishers who offer this option and can meet
the Trust's requirements
* requires electronic copies of any research papers that
have been accepted for publication in a peer-reviewed journal, and are supported
in whole or in part by Wellcome Trust funding, to be deposited into PubMed
Central (PMC) or UK PMC once established, to be made freely available as soon
as possible and in any event within six months of the journal publisher's official
date of final publication
* affirms the principle that it is the intrinsic merit of
the work, and not the title of the journal in which an author's work is published,
that should be considered in making funding decisions and awarding grants.
http://www.wellcome.ac.uk/doc_WTD002766.html
What can NSF do?
Convene federal funding agencies on policies for access to
research and its by-products created with public money.
Support the development of open source software and web service
models as an element in reducing the cost of technology particularly
in institutions that by nature have no affinity for it.
Help envision models that clearly indicate digital products
will have a chance of lasting as long as Sumerian clay tablets
and with as little effort.
Consider what role for NSF if far-reaching cyber infrastructure
issues for the humanities resolve first not in research and
teaching but around existing repositories of unique materials
that may well reveal themselves, once digitized and available
at sufficient scale, as platforms also for scientific investigation. |
|
