Data and Code for Ancient Geography: shared effort across projects and disciplines

Pleiades (http://pleiades.stoa.org) gives scholars, students
and enthusiasts worldwide the ability to use,
create and share historical geographic information about
the Greek and Roman World. Pleiades is a joint project
of three organizations: the Institute for the Study of
the Ancient World (New York University), the Ancient
World Mapping Center (University of North Carolina
at Chapel Hill) and the Stoa Consortium for Electronic
Publication in the Humanities (University of Kentucky).
Our goal is a continuously updated, authoritative digital
gazetteer for the ancient world, supporting the widest
possible range of third-party digital projects and
publications through open, standards-based interfaces.
From its earliest concept days, Pleiades was intended
to be broadly collaborative: employing, modifying and
engendering open-content information and open-source
software to accomplish its mission. This paper reports
on the associated provisions and assesses their reach and
effects within our user community, and beyond.
Pleiades employs a community-oriented, transparent editorial
process that cultivates both contributions and critiques
from the widest possible range of contributors. We
aim to echo in a virtual environment the essential processes,
workflows, resources and modes of interaction
used by scholars to advance understanding of ancient
sites, landscapes and geographic phenomena, but we
also strive to open this environment to the widest possible
range of interested participants. The bar to initial
participation is purposely set low to encourage participation
by scholars, students and enthusiasts alike, regardless
of their degree status, institutional affiliation or skill
level. All that is required is a verifiable email address and
acceptance of a contributor agreement governing issues
of professionalism, mutual respect, intellectual property,
editorial policy, assertion of identity and the open-licensing
of content. Pleiades content combines “pure” data components (e.g.,
geospatial coordinates) with the products of analysis
(e.g., toponymic variants with indicia of completeness,
degree of reconstruction and level of scholarly confidence
therein) and textual argument (e.g., comments and
reviews). In part because of this hybrid constitution—
and the varying definitions of intellectual property and
“database rights” in differing legal jurisdictions—we
have elected not to seek or assert any intellectual property
ownership in the content on behalf of the project and
its supporting institutions. Rather, our contributor agreement
assumes (and contributors must affirm) that any IP
rights inherent in the content remain with the contributors,
who grant to the project (and therefore to its users)
a Creative Commons, Attribution, Share-alike license,
which permits reuse, redistribution and remixing of the
content under clearly defined terms. These terms ensure
the widest possible range of reuse (see below), without
the need for copyright clearance requests and the like,
while guarding against trivial reorganization and restrictive
repackaging of the content that might inhibit such
reuse.
On the software front, Pleiades is entirely open source.
We make use of a number of externally developed components,
and have contributed code to some of them. The
Pleiades team has also created and released (under opensource
licenses) a number of original components. Some
of these are already in use beyond Pleiades, and we have
received contributions of code from third-party developers
for some of these.
• OpenLayers (http://openlayers.org) is the leading
open source web map toolkit. Pleiades has modestly
enhanced its features and employs it to provide contextual
maps in its web application.
• Plone (http://plone.org) is a leading open source
content management system. Pleiades has made
modest improvements to its vocabulary manager
and to its user interface framework, and contributed
these code improvements back to the Plone code
base.
• zgeo.* is a suite of Python software packages including:
zgeo.geographer, zgeo.spatialindex, zgeo.
atom, and zgeo.kml. These packages provide support
for the Pleiades Entities component and enjoy
contributions from programmers employed by The
Open Planning Project and Makina Corpus SA.
Shapely and Rtree are general purpose Python GIS software
that support the zgeo.* packages. Shapely enjoys
contributions from programmers and researchers employed
by Camptocamp SA, the University of California,
and the National Oceanographic and Atmospheric
Administration.
GEOS and SpatialIndex are low-level libraries for geometry
and spatial indexing computing. Users include
the PostGIS project and Autodesk. Pleiades has made
modest contributions to each and helped SpatialIndex
become an openly developed project.
Pleiades’ openness is driven in part by our sustainability
plan and in part by the potential for re-use. Our initial
content encompasses the compilation materials of the
Classical Atlas Project, a 12-year, 200-person international
collaboration that culminated in the publication
of the Barrington Atlas of the Greek and Roman World
(RJA Talbert, ed., Princeton, 2000). It is clear that no
academic center or institution could maintain sufficient
staff over the long-term to curate, maintain, update and
diversify this unique assemblage of geospatial coordinates,
toponymic records, temporal indicia and bibliographic
citations drawn from a wide range of specialist
literatures and primary sources. Consequently, we have
embraced the necessity of pushing out the responsibility
of (and opportunity for) creating and updating this content
to interested parties across the discipline of ancient
studies and beyond. Where these parties are employed,
professional academics, we are gambling—it is true—on
their willingness (and the willingness of their host institutions)
to absorb the redistributory costs of content
creation and maintenance as part and parcel of day-today
research, publication and scholarly communication
in their field. We are conscious that traditional “metrics”
of reward (hiring, tenure, promotion) have not yet been
adjusted to address such multi-institutional, asynchronous
and piece-wise collaboration, and so we view our
effort and our community as pioneers. Consequently, we
endeavor to surface the details of individual contribution
wherever possible: on profile pages, on individual
records, and in the change histories that underly each
such record.
The potential for reuse of Pleiades content is broad, and
we hope to see both unexpected and serendipitous reuse
cases arise from outside the project team. Our early collaborators
are interested in a range of useful applications
that can be foreseen or are already in prototype. One
chief class of use is as an “authority list” for Greek and
Roman geographic names and the locations associated
with them. Both existing and new databases and digital
resources can make use of Pleiades content (and our
stable URLs for discrete elements therein) to refer unambiguously
to the places and spaces mentioned in ancient
texts, the subjects of modern scholarly works, the minting locations of coins, and the findspots of inscriptions,
papyri, and the like. Using Pleiades as a central
geographic authority reduces opportunities for ambiguity
(consider that we know of 19 distinct cities named
Apollonia in antiquity), while setting up the possibility
of cross-project services and data sharing that exploits
the common standard. In addition to the human-readable
HTML interface, Pleiades provides access to its content
in simple, standard formats that can be harvested, or aggregated
dynamically, to produce dynamic maps using
third party tools and services like Google Earth or Yahoo!
Maps. We are also working with other projects to
develop standards-based mechanisms for cross-project
geographic search (e.g., relevant information within
30km of a named place). These same functional components
will make it possible for scholars and students
alike to pull Pleiades content into their own research and
teaching tools and contexts, using them to solve problems,
explore possibilities and produce map visualizations (mash-ups) for further sharing, reuse and publication.