Library Collaboration with Large Digital Humanities Projects

At DH2007 a special session on “finishing” large humanities
research projects (now forthcoming as a
cluster in DHQ) suggested, in part, that particular stages
of such projects might be completed, but that continuing
institutional support was important for the longterm
sustainability of the projects and their products. At
DH2008 a special session was devoted to “Aspects of
Sustainability in Digital Humanities,” in which technical,
organizational, and scholarly dimensions were discussed
with reference to a museum project, along with
metadata and the issue of portability in other settings.
In this paper, we would like to continue the theme of
sustainability. We will discuss issues of institutional
support for a large digital humanities project, and then
propose collaboration with the university library as the
only realistic option for long-term sustainability in our
environment. We believe that our experience is typical
of the situation for other projects, large and small, that
many digital humanities faculty now face at their institutions,
and therefore that our experience is also typical of
the demands that will be placed on libraries to sustain
faculty digital research for the long-term.
As for many digital projects, the Linguistic Atlas Project
(LAP) began with computing resources located within
the research office itself, first personal computers and
later servers. When the university created a research
computing service (as an addition to the institution’s
administrative and instructional services), LAP was one
of the first clients--the editor of LAP was even asked
to help design the service. However, over the course
of several years the funding structure for the research
computing service changed from an essentially institutional
budget with additions from externally funded research,
to a fee-based service much more dependent on
research with annual external funding. This meant that
humanities projects like LAP, while not excluded from
the research computing service, either needed to find
consistent external funding or hope for sufferance from
the paying customers. Neither of these options appeared the long term, and, citing budget pressures, the university
administration declined to guarantee funding for gaps
in the external funding that LAP has regularly solicited
and received. LAP reacquired a server for its research office
and 20 Tb of storage space for its large and growing
multimedia archive.
Enter the University Librarian. During discussions
about institutional support for LAP, whose paper and
audio tape records are stored in a library special collections
facility, the University Librarian suggested that the
library was expanding its multimedia collections (e.g.,
our institution hosts the Peabody awards, and so has an
important interest in archiving TV materials), and even
the large multimedia collections of LAP could be incorporated
there. Unlike the research computing service, the
mission of the library includes archiving and dissemination,
now increasingly of digital materials as well as traditional
paper. We are now developing cooperation not
just for a multimedia archive, but also for dissemination
of LAP products and information through library means
in the context of continuing scholarly activity.
The terms of our cooperation can be discussed under just
the headings of the DH2008 sustainability session:
—Technology. Two issues dominate here, long-term
storage and security. It had been planned that LAP materials
would be stored on spinning disk in the research
computing service storage array. However, the library
holds its multimedia collections on LTO-4 tape. The
latter is much more cost effective; however, tape storage
requires periodic refresh of the medium (which the
library will carry out indefinitely as part of its archival
mission), and limits the speed of access to the materials.
LAP and the library intend to disseminate the project’s
multimedia (audio, and later image) products, but the
speed of online access is not required for all materials.
We will discuss what can be disseminated online and by
other means and the terms of such dissemination, as well
as the balance between spinning disk and tape in a setting
like ours.
Security is an issue for the online presence of LAP at the
library. The project has had a heavily interactive Web
presence including GIS for over a decade, first on servers
in the research office and later at the research computing
service. However, the library has not provided interactive
Web access of this kind before. We will discuss
integration of LAP programming into the library server
system, including reversal of the previous practice of
heavy server-side scripting to place more emphasis on
client-side processing.
—Organization. LAP and the library will use metadata
standards developed for library and linguistic collections,
but the question of standards is not trivial because
metadata practices for libraries and linguists have not developed
in parallel owing to different goals of the practices.
We will discuss the resolution of these differences.
We will also discuss the organization of the partnership
of the library and the LAP. Just as for the research
computing service, there are real costs associated with
continuing services provided to LAP by the library.
However, the difference in the mission of the research
computing service as compared to the library makes a
difference in how such costs might be borne over the
long term. Periodic infusions of external funding can assist
the library to accomplish its goals, but the improved
match with LAP as an archival and information resource
makes it more reasonable to provide continuing support
for the project without annual external funding.
—Scholarship. LAP cannot just be given over to the
library in its entirety because it remains very active in
scholarly work. As long as LAP continues to conduct
new research, it can cooperate with the library to keep
its Web presence up to date including both links and
programming for interactivity. The LAP archives can
be extended with new material, but also with new versions
of old materials (such as digital images of paper
records). We will discuss how new work and additional
aspects of older materials can continue to round out LAP
collections in the library, as an active partnership instead
of a static archive.
We believe that cooperation at our institution between
the library and a large digital humanities project should
not be a singular occurrence. Some digital humanities
projects have always been associated with and supported
by the university library, and many libraries, including
ours, have been active with their own digital initiatives.
However, a great many more digital humanities projects
began as separate faculty initiatives, and many libraries
are now interested in developing their digital archives
and activities. If independent projects are to be sustained
beyond their initial development, they will require the
sort of new partnership that we discuss. Libraries, too,
can enhance their digital activities through cooperation
with faculty digital humanities projects as information
resources. We hope our partnership can provide a model.