New World Ordering

In April of 2008, LLC published a thorough survey by
Martyn Jessop of many factors contributing to an “inhibition”
of the use of GIS, or Geographical Information
Science, in the digital humanities (Jessop 2008). That
GIS has been slow to penetrate a scholarly population
generally receptive to new practices and technologies
begs a discussion of issues at once historical and methodological,
institutional and pragmatic. It also demands serious
engagement by scholars, programmers, librarians,
and advocates for shared data and transparent, flexible
services. To be effective, this engagement must come at
many levels simultaneously: we must work to build core
infrastructure to support GIS and leverage the strengths
of (primarily government and academic) data providers;
we must carefully analyze past successes as well as failures
in the digital humanities in order to move forward
with more robustly-imagined scholarly projects; and we
must interrogate both a toolset that has evolved to suit
scientific inquiry (that is, positivist models of physical
behavior and dense, detailed, precisely-defined data sets,
generally synchronic) and our own inherited systems for
interpreting the human record within a spatial field, georeferenced
or conceptual. Above all, we must make a
concerted effort at understanding what it is we do, when
we “do GIS.” This panel will provide differing perspectives
on GIS for humanities scholarship, but from within
the coherent narrative of a University of Virginia Library
effort to build modern infrastructure, support innovative and conditions of our community’s strange inhibition.
UVA Library’s Director of Digital Research & Scholarship,
Bethany Nowviskie, will open this discussion, but
in order to place it within a real-world services framework,
Scholars’ Lab GIS Specialist Kelly Johnston and
programmer Adam Soroka will describe our efforts in
building an infrastructure that includes sophisticated
hardware and several layers of software: a datastore for
standards-compliant metadata, vector, and raster information;
a translation layer that provides Open Geospatial
Constortium (OGC) services to GIS applications and
directly to the Web; and an application layer including a
discovery portal and browser-based, interactive mapping
systems that consume OGC web services.
For much of its history, GIS work has been predicated on
a model of powerful independent workstations featuring
complicated monolithic closed-source software, large
locally-stored datasets, and users working in relative isolation
who publish results in traditional media (Lo and
Yeung). Until recently, this has been the model in play at
the University of Virginia Library and at most scholarly
institutions. A new way of working, introduced by the
OGC and gaining increasing traction, utilizes Web services
such as Web Map Service (WMS) and Web Feature
Service (WFS) and creates an environment where processing
occurs server-side, data is stored centrally and
shared, and both processes and results can be published
digitally, in networked media. The UVA Library Scholars’
Lab brought this new way of working to our local
teaching and research community through a GIS spatial
data infrastructure project (colloquially called the “New
World Order”), in continued development since 2008.
Its tools enable new forms of collaboration, new forms
of publishing, and new forms of pedagogy – and they enforce
better internal stewardship of GIS datasets through
standardized metadata.
As is common, the UVA Library geospatial data collection
has grown episodically, in many formats over many
years, in response to user need and data availability. In
2007 we resolved to find a strategic solution to streamline
user access to datasets and improve our ability to respond
to user requests. Johnston will review the criteria
used in our selection process and explain our software
choices. He will address: how this new infrastructure
better supports collaboration among scholars building
spatial datasets across projects and disciplines; how it
undergirds the Library’s efforts in classroom support by
simplifying the contribution and sharing of datasets for
collaborative pedagogical activities; and how we identified
standard metadata as a key to success and embarked on work to improve both the quantity and quality of our
spatial dataset.
To illuminate the advantages of this approach from a
developer’s standpoint, Scholars’ Lab GIS Specialist
Christopher Gist will offer a case study in digital scholarship
using GIS technologies, new and old. In 2005,
UVA Library’s GeoStat Center (now part of the Scholars’
Lab), in cooperation with the Virginia Center for Digital
History, developed a Web-based student collaboration
tool for a history class being taught by Professor Ed Ayers.
The Southern History Database (SHD) aimed to
allow students to pool their research on primary sources
to provide larger datasets than would be possible for individual
students to collect. The process was also meant
to be iterative, so that one class might build on the work
of previous semesters.
The SHD had a large spatial component, which involved
students’ determining the location of events referenced
in their primary sources. These results were then tallied
and presented in thematic map form. Users could
click the map to retrieve results for a given location.
This portion of the project presented some very difficult
technical issues for the development group. Traditional
web mapping applications require specialized server applications
and incurred great expense in licensing and
management. Available open source applications were
limited to static data.
Gist will describe how the development team cobbled together
several applications, including MapServer, using
the scripting language Perl to create a custom map file on
user request. This method is server intensive, difficult
to maintain long-term, and not scalable. If new features
were to be added, existing scripts must be completely
rewritten.
Since that time, many new open source applications
have risen from the OGC movement to standardize spatial
services and applications, and are available through
the Open Source Geospatial Foundation. Since all these
tools use common standards – the same ones on which
we are building our GIS infrastructure at UVA – it has
become easy to develop mapping applications that allow
for the creation, storage, management, delivery and
display of spatial data.
Now instead of large and expensive proprietary spatial
web applications, the paradigm has moved to lighter
tools that can be used together through OGC standards.
This will allow us nimbly to create complex applications
like the SHD in a matter of days instead of weeks
or months. And unlike one-off applications, these new many formats that can be used in Google Earth and other
easily-acquired tools. All these features make spatial
applications for humanities scholarship much more attainable.
But will they help us overcome our essential inhibitions
when it comes to GIS in the digital humanities?
These locate themselves not only, as Jessop reminds us,
in questions about the nature of the tools and our facility
with them, but also in more theoretical and methodological
concerns surrounding the status of image-based
scholarship and our critical stance toward space and
place. Bethany Nowviskie and UVA Scholars’ Lab coordinator
Joe Gilbert continue the discussion with a look
at the exigencies of humanities information and our critical
impulses.
How do we conceive of and create spatial tools that are
neither strictly mimetic nor strictly symbolic? One obstacle
we may encounter is that maps, particularly in
their digital embodiments, primarily seek to represent
real-world spaces and objects. We are therefore limited
by these tools’ seeming inability to engage objects on
multiple semiotic levels. If we can only use spatial tools
to investigate what C.S. Peirce termed “iconic,” or representative,
types of signs, we are artificially restricting
the kinds of interpretive acts with which we can engage
(Peirce). Current GIS tools elide the subjective nature
of both time and space, and thus the possibility of using
such tools as richly signifying spaces in themselves.
Nowviskie will offer a young lady’s 1823 “Book of Penmanship”
from the David Rumsey Historical Map collection
as an example of a sophisticated (if naïve) humanities
GIS—an historical and artistic document that
embeds a brand of subjectivity that the tools and frameworks
we build must be able to accommodate (Henshaw).
This artifact—with its fascinating imaginative
reconstructions, in text and image, of 1820s American
geography—will also be presented, more polemically, as
the physical embodiment of a method that we must take
seriously in digital humanities scholarship if GIS is to
take hold: graphesis, the visual construction of knowledge.
Gilbert will conclude with a look at a spatial signifier
that has become iconic in a completely different sense
– the Google Map marker. Here we encounter another
potential difficulty for spatial analysis in the humanities:
the symbolic language used by prevalent tools is too divorced
(to paraphrase Raymond Carver) from what we
talk about when we talk about the world. Here, the fear
of maps as representations of the real is avoided, but is
also replaced by an empty, literally (or virtually) floating
set of signifiers. Rather than depicting this limited sign
system laid over the rigid lattice of latitude and longitude,
can we re-think Google Earth’s virtual sphere of
photorealism and arbitrary markers through the lens of
Yuri Lotman’s concept of the semiosphere, a “synchronic
semiotic space?” (Lotman). A future avenue for investigation
implicit in this argument would be an understanding
of mapping interfaces in terms of Baudrillard’s
notion of simulacra.
References
Henshaw, Frances A. “Frances A. Henshaw’s Book of
Penmanship Executed at the Middlebury Female Academy.”
April 29, 1823. In the collection of David Rumsey.
Pub list no. 2501.000. http://davidrumsey.com/
Jessop, Martyn. “The Inhibition of Geographic Information
in Digital Humanities Scholarship.” Literary and
Linguistic Computing, vol. 23 n. 1; April 2008.
Lo, C.P. and Yeung, Albert K.W., Concepts and Techniques
of Geographic Information Systems. New Jersey:
Pearson Prentice Hall. 2007.
Lotman, Yuri. Universe of the Mind: A Semiotic Theory
of Culture. Bloomington: Indiana UP. 2001.
Peirce, C.S. Writings of Charles S. Peirce: a chronological
edition. Ed. Fisch, Max H. Bloomington: Indiana
UP. 1982.