ARLIS/NA
30th / VRA 20th
Plenary Session II
The
Three Vs:
Moderator:
Katharine Martinez,
Harvard University
Speakers:
Jennifer Trant, AMICO
(Art Museum Image Consortium)
Max Marmor, ArtSTOR,
The Andrew W. Mellon Foundation
Susan Jane Williams,
Art Library, Yale University
Paula Lupkin, School
of Architecture, Washington University
Recorder:
Rebecca Price,
University of Michigan at Ann Arbor
Sponsor:
Rice University,
under the auspices of the Vice President and CIO and the Executive Director of
the Digital Library Initiative.
The session began with a welcome from Katharine Martinez. She introduced the session saying that the four presentations would together provide an overview of current directions in visual technology, visual culture, and visual literacy. Martinez noted that the themes were building on many discussions at the conference, looking particularly at users, uses, and collection development of digital image collections.
The
format of the session was such that each speaker spoke for ten minutes
presenting his or her view of the current state of visual technology, visual
culture, and visual literacy. This was followed by approximately thirty minutes of
questions posed by Katharine Martinez and then thirty minutes of questions and
comments from the audience. This
report will follow the same format with each presentation summarized followed by
the questions with their responses noted.
[Note
from the recorder: the style of
each talk varied and this summarization will reflect those various approaches.
I apologize ahead of time if I have misrepresented any speaker's
intent with my summaries.]
Jennifer Trant, Executive Director, AMICO
Trant introduced her remarks by saying that rather than talking about AMICO, she would instead focus on her own research experiences as an art historian and historian. Noting that the art historical canon is relatively small, she made the point that barriers to the study of images are related more to institutional and discipline related structures than to content. She would like to see products that better enable interdisciplinary research. The gist is that one can always add content over time (and content will inevitably change over time), but that throughout the process attention needs to be given to the description and presentation of that content.
To
“enable interdisciplinary research” raises two important issues for Trant:
(1) access and (2) cost.
Trant
stressed that to achieve success a revolution was not required, but that she
feels there needs to be an evolution— an incremental approach—to change, the
point being that there are foundations of use and uses in place.
The user, both traditional and new, needs to be brought into the dialogue
as we adapt and develop digital technologies to meet existing and potential
needs. At the same time she cautions that we should think carefully
about the point at which users need to be involved in the conversation.
They do not necessarily gain from being part of the complex discussions.
What
are these foundations on which to build and create an image network?
Issues relating to the digitization of cultural heritage materials have
been addressed at the Getty, which led to the MESL (Museum Educational Site
Licensing) project [see Delivering Digital
Images. Cultural Heritage Resources for Education (MESL, vol. 1) and
Images Online. Perspectives on the Museum Educational Site Licensing Project
(MESL, vol. 2], both edited by Christie Stephenson and Patricia McClung, J. Paul
Getty Trust, 1998). The MESL
Project allowed the testing of concepts and practices by several museum and
academic institutions. These early
projects produced valuable conclusions, such as:
There is value to having museum collections available digitally.
There are collaborative opportunities in this work (in fact collaboration
should be encouraged in order to discourage redundant work).
They provided a basis for understanding and establishing technological
needs.
Trant
concluded her remarks by urging a positive spin.
She argued that we should see contracts as empowering, not restrictive.
She cited the example of AMICO as providing an economic model, a model in
which the rights issues are faced head-on (especially in regard to contemporary
art), a forum for museum collaboration, an incentive for local investment in
digitization, and an important means of building bridges between collections and
users.
Max Marmor, Director of Collection Development, ArtSTOR
Marmor's
rapid-fire presentation lends itself to a bulleted recap, which is reflected
here.
Aims of ArtSTOR
Provide a user-centered digital library for integrating digital images
into teaching, learning, and scholarship
Provide a regulated, scholarly space ("regulated" referring to
the goal of creating on the Internet a "protected environment" for
education and scholarly use of digital images).
Lessons from the
model of JSTOR for this digital library system
Scalability (build an infrastructure capable of sustaining ongoing
collection growth)
Sustainability (build in mechanisms for progressive cost recovery, as
compatible with a non-profit enterprise that seeks to be as affordable as is
feasible)
Dependability (build a robust, reliable service—both in terms of
content, data, and ongoing availability)
Why is Mellon funding
this?
The primary mission of the Andrew W. Mellon Foundation is to serve
museums and institutions of higher education. (See <http://www.mellon.org>
for more information regarding the foundation.)
Mellon hopes to provide collective solutions to collective problems and
this is inarguably a collective problem.
There is a high cost associated with developing the infrastructure
necessary for meeting the collective needs of these institutions; local
resources are not adequate to creating such a public utility.
What will be offered
on ArtSTOR?
Collections that will support, in the very near term, typical
undergraduate art-historical teaching needs.
At this point ArtSTOR envisions a starting offering of approximately
250,000 images.
The breadth of the collections—ranging from "teaching
collections" to scholarly collections of primary source materials—will
reflect their work with museums, research projects, publishers, and photo
archives to develop content.
Awareness and use of
metadata
ArtSTOR will comply with community-based standards.
By necessity metadata will vary somewhat from collection to collection.
Mellon is very interested in metadata harvesting and sharing and looks
forward to working with interested institutions on this front.
How is ArtSTOR
dealing with image copyright?
Images are intended only for non-commercial educational use and cultural
purposes
ArtSTOR will create a "protected environment" within which
intellectual property rights will be managed appropriately while educational and
scholarly activities will be strongly supported.
ArtSTOR seeks non-exclusive, perpetual rights with its image suppliers,
perpetuity being essential if end users are to plan confidently on the ongoing
availability of the service and collections.
This is admittedly a complex, changing environment and demands will
change over time.
What is the timeline
for ArtSTOR?
In 2002/03 preliminary "testing" at a few institutions will
begin.
Depending on lessons learned during this testing period, a service might
be launched in the course of 2003.
For more information,
there will be a website:
Susan Jane Williams, Visual Resources Curator, Art Library, Yale University
With
a little help from Brittany Spears, Williams offered what she called a
"view from the trenches." Brittany's "Oops, I did it again," provided a comic
and poignant illustration of our current practices of spending money and time on
repeated efforts across collections and of spending money on poor quality
digital images.
There
is a cost to doing it wrong. Williams
urged us to think about routing our digital interest into rational, collective
collection development. There are
over eight hundred visual resources collections in the United States.
Most are in academic departments rather than libraries. All use local classification/cataloging schema.
With the era of digital images, there is a compelling need to work toward
greater use of standards across collections and a desperate need to build
coalitions at local, regional, and national levels.
Our emphases should be on reducing redundancies, on setting local
priorities with local, unique materials, and on collaborating on shared needs. This boils down to putting our energies into developing the
infrastructure, not the product. Once
the infrastructure is in place, the product will follow.
Williams
offered a list of rules to work by to help keep on the straight and narrow:
Do not get locked into HTML. That
is, don't spend time creating stand-alone, limited-use webpages.
Instead, work on data structures/metadata that will serve over time.
Avoid quick and dirty solutions. Think
long-term.
Avoid just-in-time work (meeting immediate needs rather than long-term
needs). Admittedly, given faculty
and departmental demands, much of our work may be "just in time".
It is our responsibility to speak up and alert our faculty to the
wasteful nature of that strategy.
Work to share records. Ideally
work toward using OAI (Open Archive Initiative) to share metadata across
collections. (For more information
on OAI, see <http://www.openarchives.org>.)
While building image systems, think of these tools as necessary to any useful image delivery system:
Search and discovery. Develop robust search and browse systems for finding images
using any number of access points.
Ability to save groups of images.
This is based on the
model of the study gallery, but any number of uses and applications could
develop from this capability.
Presentation preparation tool.
Here the model is the light table, but it will be interesting to see how
the online world can expand that two-dimensional model to more complex
possibilities.
Large image file handling.
Work toward integrating with other
media types (e.g., sound). In
addition, be aware of other new possibilities that are available.
Funding
is always of primary concern. Because
of this and the realization that time is money, Williams seconded Marmor's
earlier statement of the need to build complex systems over time, and in stages. Realizing that the “perfect” online image delivery system
is not going to happen overnight or in one place, scalability has to be built
in.
Inevitably
there are local, yet shared, challenges we face as we steer our vessels in this
new direction [pardon the recorder's metaphor].
Be aware of these issues, expect them, and meet them head on.
Last minute requests
Incorporation of private images
User authentication
Williams asked the perennial (and politically-charged question), “Who should do the work?”
She argued that libraries, which have been doing this sort of thing for
years, should provide the model.
Local cultures vary, but in most cases collaborations with the local
Information Technology group or a partnership with the IT group and the library
may offer valuable working models.
Williams concluded that our greatest current need is to agree on metadata
standards. With these in place, we can start the work of creating data
systems that can be shared, thereby reducing the need to say, "oops, I did
it again."
Paula Lupkin, Assistant Professor of Architectural History, Washington University School of Architecture
Professor
Lupkin, who teaches architectural history at Washington University, opened her
remarks by saying that she was speaking as a “user” of digital images.
She uses digital images in the classroom, in course webpages, and in
presentations. She shared some of
the problems that she currently has with online images, showed us some of the
images she uses in teaching, and expressed her desires for a more useful and
useable resource.
Her
main point boiled down to a frustration with the lack of flexibility, control,
and choice in images. This
frustration was magnified by the contrast of these inadequacies with the
flexibility of the canon and of the technology.
She pointed out that few professors continue to teach using the old
standard chronological history. Instead,
history teaching is approached episodically and thematically.
Such a deconstructivist approach requires that collections be fluid,
dynamic, and responsive.
Lupkin
teaches the survey course from a global perspective (not just Western
architecture) to meet new standards of NAAB.
She has had to develop both new content and a new intellectual framework,
but has found it very difficult to find images to support this new content and
new approach. For instance, she
uses an approach adopted by world historians, in which the historical
investigation is on the emergence of global interactions.
Rather than focusing on a particular culture and location, Lupkin takes a
macro-level view and focuses on the interactions between cultures and
civilizations. The emphasis is on
teaching process rather than product. The result is a view that is not comprehensive but is,
instead, selective with chosen and specific times, locations, and themes (e.g.,
trade, migration, and conflict). Our
image collections need to have access points that allow extraction into dynamic
arrangements.
Despite
working toward this “new history," Lupkin finds that her course is still
too Western in focus, as it is difficult to find enough Eastern images to
balance coverage. Adding new images is time-consuming, if they can be found at
all.
In
conclusion, Lupkin argues that we need to keep and support a plurality of
approaches. Old models do not work,
but instead we need to develop systems and resources where many, many images are
available in diverse ways for any number of teaching approaches.
Lupkin's thoughts and frustrations shocked none of us, as she echoes many
of the troubles and complaints we hear from our own faculty and users. She does provide a vivid reminder that we need to be aware
not simply of technological changes and opportunities, but changes and
developments in content needs.
Questions
from Katharine Martinez:
JT = Jennifer Trant
MM = Max Marmor
SJW = Susan Jane
Williams
PL = Paula Lupkin
KM = Katharine
Martinez
1.
Have you implemented or participated in a study or survey about the use
of images? What did you learn about
users' needs and what is their influence in shaping content?
JT:
AMICO continues to be very interested in connecting the people creating
the content with those using it. They
have tested it with focus groups (AMICO testbed and public school testbed).
She noted that identifying how users want to use the collection is easier
than collection development. They
also did a CAA focus group where they received the not surprising message that
faculty want all images digitally. Of
course, Trant points out that you cannot have everything at once:
you can focus on access or you can focus on collection development.
We need new, contemporary collections, not just retrospective
collections.
Access
will inevitably change teaching methods and at this point, change is more often
led by students than by faculty. The
use is unpredictable. One can plan
using past models of use, but one has to be ready to adapt quickly as patterns
of use change.
SJW:
Have to know that statistics of study are valid and that the survey is a
good instrument. Assessment is a requirement.
MM:
Marmor noted that he was involved in a survey done at Yale by the Kress
fellow last spring (look at VRA-L archives for more information).
Users want both Western and Eastern content.
There is still a need to replicate analog and digital.
At Mellon the studies show that everyone wants everything (in other
words, no concrete results to help prioritize efforts).
2.
How would you advise the audience about making collection development
decisions? In addition, what
technological issues must the visual resources professional consider as more
licensed products and Web-based resources become available? What do they need to know before providing such a buffet of
image databases to users?
SJW:
Put efforts into developing local projects, that is, projects rich with
local content and meeting local needs.
PL:
Advises that users do not want small, limited collections.
SJW:
Explains that the smaller collection approach is a block-building
approach in which collections are built on collections—collections are then
shared and made accessible as broadly as possible.
[With smaller collections agreeing on metadata standards and using OAI,
those collections join together virtually to provide a single large collection.]
MM:
Notes that we need to be careful about managing expectations regarding
the extent of any given collection. ArtSTOR
is hoping to help put an end to redundancies by providing those "staples of
art history teaching." Local
collections should work to meet those pedagogical idiosyncrasies unique to each
institution.
3.
How can I plan—how can I budget? How
do I respond to faculty demands to have images now?
JT:
There is no one source for material.
One needs to plug into the institutional structure to meet the demands of
a multi-disciplinary user group. As
you plan, plan for scale, diversity, and both non-traditional and traditional
users. There is power in moving the
collection out of the art history department and into a campus-wide resource for
visual literacy and visual communication.
SJW:
Responds that the other side to managing a budget is "managing
expectations." In fact, that
is part of the problem. You need to
show faculty what is great about digital—not PowerPoint—networks.
We need to be proactive in taking the message to faculty.
PL:
Asks what do you mean by network?
SJW:
The source for images should be a (shared, collaborative) network, not a
local (limited, restricted) source, because you just cannot do everything
locally.
MM:
Cautions that it is not so much about building digital collections as it
is about organizational change. The
situation calls for conversations across boundaries.
The challenging thing is how to foster those conversations.
Questions/Comments
from the Audience
Joy Kestenbaum (Pratt
Institute):
Points
out the conflicts arising from our efforts to meet all needs.
Need to pull things from everywhere.
Cannot focus on one thing or one source. Needs to be user oriented more than collection oriented.
SJW:
Williams reinforces her message of the need to create sustainable
collections--not HTML webpages.
PL:
Lupkin sees that her mission is to publicize the issues and the needs to
fellow faculty.
MM: (as footnote to
Lupkin's comment) Angela Miller, University of Washington, is doing similar
things in American studies (that is, talking and teaching across disciplines).
Michelle Murray
(Sarah Lawrence College):
Wants
to make three observations:
1.
She notes an inherent conflict in which VR managers talk about pooled
resources and faculty/departments talk about hoarding resources (focus on
ownership).
2.
As collection managers we look at local constituency.
She thinks that we need to think more broadly.
Think about what local resources are of broad interest.
3.
Noting that there are hierarchical distinctions and that, generally, faculty are
not receptive to VR curators, she wonders who will get the message across to
faculty?
KM:
Martinez notes the importance of collaboration between VR professionals
and IT (Information Technology) and library staff.
JT:
Observes that teaching and research are very personal endeavors.
We need tools for each teacher to develop his or her own courses and to
enable individual interpretations. The
campus needs to work together to provide resources (library, VR collections,
departmental archives, etc.) and the faculty need to be involved in the creation
of the content and the resource to ensure that teaching needs are being served.
Nensi Brailo
(California College of Arts & Crafts):
A
question for Max Marmor and Jennifer Trant: She asks why collections are being
duplicated—would it not make sense to create unique, shared collections and
not duplicate efforts?
Jennifer Johnson
(Occidental College Slide Library):
Johnson
notes the importance of Lupkin's talk and examples showing that images come from
so many sources and that only about 20 percent of the images were available from
a vendor. Maps, didactics, and
diagrams are all the result of in-house production.
Will ArtSTOR be able to provide such types of materials?
MM:
Suggests that we need to think in the long term.
At this point, he hopes that ArtSTOR may be able to provide 20 percent of
needs, perhaps in five years it can be meeting 50 percent of needs.
He cautions that one can never anticipate all needs and that local and
licensed collections will need to be integrated.
KM:
Notes that the challenge is to involve other campus groups (particularly
for technical help). The value of
working with the VRA core enables integration of collections.
SJW:
Suggests that the library OPAC is a useful paradigm.
Local content is always being added.
Cataloging is always the most expensive part of the process, and that
part needs to be shared as much as possible.
JT:
Trant remarks that she is seeing more sharing of locally created
material. This raises a
question for her: What interpretive
and descriptive materials are used and/or needed?
Eric Schwab (Cornell
University):
Notes that standardization is the keystone of our work. Asks if organizations like AMICO and ArtSTOR see image delivery as standard as cataloging? He would like to deliver through one system.
MM:
Yes, ArtSTOR will be following metadata standards.
He notes that the question is about inter-operability, which does need
much more thought and attention for the issues to be recognized and resolved.
KM:
Hopes that future conferences will have sessions on models of image
delivery systems in order to address such issues.
Allan
Kohl (Minneapolis College of Art & Design and AICT):
Asks
about the possibility of shared cataloging—notes that it is a very intriguing
idea and ants to know how his building block can be part of the larger
construction. How can we facilitate
the sharing of data?
SJW:
Enthused about Open Archives Initiative (OAI), which sets a relatively
low technology bar. {More information can be found at <http://www.openarchives.org/>.]
We could write XML
schema for the VRA core and the data would then be harvested using a mapping
document. The data could then be
fed directly into the local catalog.
MM:
Before his remark, Marmor acknowledges the important role that Allan has
played by developing AICT (Art Images for College Teaching at <http://www.mcad.edu/
AICT/>). Suggests that Mellon
may have a role in taking on some of the risk of shared cataloging.
He repeats their interest in metadata harvesting. The scalability and
dependability of the system need to be investigated.
JT:
Trant cautions using a technical standard to solve an intellectual
problem. Standards are only as good
as the level at which they are implemented.
We need to talk about where we need 100 percent compliance and where we
can get away with less than 100 percent compliance.
Christine Sundt
(University of Oregon):
Is
struck by how this seems like new technology in an old wine bottle—feels we
are trying to fit this technological transition into a nineteenth- century
pedagogical framework. We need to
see how to teach different courses and different ways of teaching.
She sees that the challenge is to serve all.
Henry Pisciotta
(Pennsylvania State University):
He
sees the struggle of getting content and of understanding what that content will
be surfacing as the primary consideration.
He asks the panel for speculation on technological tools to help identify
content that will be needed. What
tools are out there to know how to find or to know what users need?
KM:
Based on existing information and practice, users want diverse
images—from comic books to painting to architecture.
We need to remind ourselves that libraries were not built in a day
either. Researchers go on a holy
pilgrimage; the physicality of traveling, of browsing, that creates
understanding and scholarship. We
need to pat ourselves on our backs and realize that we are here working together
on national solutions.
JT:
An observation: The previous week Trant spent three days working on national
and international standards for cultural needs.
She thinks that the angst over the scope of and access to image
collections is our collective discomfort over the pace of change.
She reiterates a common theme: we need to look at the long term.
Henry Pisciotta
(follow-up comment):
Even
if everything were digitized, one would need still more.