Roundtable Panel: Modeling and Visualizing Historical Narrative

The modeling and visualization of temporal
phenomena—events and narratives—is an important area
of research in the digital humanities that has received relatively
little attention to date. There are some exceptions. Matt Jensen’s
SemTime, Johanna Drucker and Bethany Nowviskie’s Temporal
Modeling Project, the Center for History and New Media’s
Timeline Builder, and Bruce Robertson’s Historical Event
Markup Language are important projects that have been
presented to or developed within this community. However, in
spite of these exceptions, the limits of work in this area are
clear when compared, for instance, with the closely allied area
of spatial modeling and historical GIS. In recent years,
interactive digital maps have become increasingly prevalent.
In addition to the ubiquitous Google Earth, other map resources
produced by universities and textbook publishers allow users
to pan and zoom, hyperlink to related content, control layers
of spatial information, and select temporal ranges. However,
history textbooks are filled with timelines as well as maps, and
temporal thinking is critical not only to history, but to many
other disciplines besides. Nevertheless, in contrast to mapping,
the problem of modeling and visualizing time in an interactive
and digital environment has received almost no attention. With the hope of raising the visibility of current efforts and
inspiring new research, this panel brings together six individuals
working in the area of temporal modeling and interactive
timeline development. Our goal is to introduce current
developments; to identify interesting and significant areas of
development that will make temporal modeling and
visualization into a robust field of research, and to discuss
models for an interactive database and historical event
visualization system.
Some components of a historical event model are well
developed. Temporal ontologies such as DAML-Time and ISO
19108 offer essential guidance for formalizing temporal objects.
The Historical Event Markup Language and, more recently,
the Named Time Period Directory Standard have developed
XML models for describing historical events. Creative efforts
to improve timeline visualization are surprisingly limited, but
there are several noteworthy exemplars. The MIT Media Lab’s
SIMILE project has an intuitive and flexible interface and
allows several timelines at different scales to be manipulated
together. However, it implicitly embodies a historical model
of individual unambiguous and unrelated events rather than
complex narratives, replicating the “tickmarks on a line” model
familiar from static textbook and wall chart timelines. By
contrast, the Temporal Modeling Project and SemTime are
visualization experiments that allow developers to incorporate
agency, causality and relationships.
Better handling of relationships is among most important issues
for future development of temporal modeling in the digital
humanities. Historical events have complex, multiple and
perspectivally unique relationships to one another. Historical
events and narratives also all have a relationship to spatial
information. Events occur somewhere, even when they are
global, and the same event (the Neolithic, modernity) may occur
at quite different times in different places. Finally, timelines
and event databases can be considered as components of holistic
information systems. All of these issues will be explored in this
roundtable session. Other topics that have not been addressed
by previous researchers in temporal modeling include the
representation of temporal and temporo-spatial ambiguity and
uncertainty, and the specifications for historical event systems
and services with multiple users. These topics, too, will feature
in our planned roundtable. All of these areas have implications
for modeling of both data and systems, and also for very new
kinds of visual representation.
The proposed Modeling and Visualizing Historical Narrative
Roundtable brings together a number of individuals who are
working actively in this area. This panel includes some
individuals (Drucker and Robertson) who have been researching
temporal modeling and visualization for some time, and others
(Johnson, Lancaster, Mostern, all affiliated with the Electronic
Cultural Atlas Initiative) who have worked primarily on spatial
visualization for history and the humanities and have begun to
engage in research on temporality as an extension of that
interest.
Ruth Mostern, as panel organizer, will introduce the panel with
a discussion of work-to-date on temporal modeling and possible
directions for future research. The rest of the panelists will
discuss their own areas of development, as follows:
• Ian Johnson will discuss developments in the Heurist
generic collaborative content creation system and the
TimeMap spatial browser to create a model and visualizer
for historical events, illustrated through the example of a
Silk Road timeline project. He will discuss the underlying
data structures, planned developments in the data model
used to describe historical events, and research on improved
visual methods of entering and viewing the web of
relationships between historical events, including timeline
visualizations.
• Lewis Lancaster will demonstrate the What, Where, When
and Who prototype for library catalogue searching and web
browsing. In this system, a directory of named events
extracted from Library of Congress subject headings serves
as the basis for structured searching. Searchers can browse
a library catalogue by using a timeline and a map linked to
the time period directory and a place-name gazetteer rather
than conducting a traditional text-based search. The user
may expand the search from a library catalogue to include
Wikipedia and other on-line systems. Named events, place
names, and personal names are interrelated to allow complex
search parameters.
• Bruce Robertson will talk about the latest developments for
the Historical Event Markup Language. The first of these
is an RDF syntax for historical events that makes historical
markup of timed, labeled events associated with persons,
places and keywords simple for students and others. The
other development is the use of OWL web ontology
reasoning to express chronological reasoning (i.e. how we
know that the battle of Marathon was in 490 BC). This will
be associated with RDF reification
• Johanna Drucker, an important researcher in this area,
regrets that she will not able to attend the conference due
to a prior commitment. However, she has offered to produce
visualizations based on her recent work to be presented by
the conference organizer.
Because of the large number of presenters and the desirability
of open dialogue, we are proposing a roundtable format. The
panel organizer and each presenter (including Drucker in
absentia) will make a presentation of 10 minutes. Each presenter
will discuss his or her current work and future plans in the area
of temporal modeling and visualization. This will allow time
for a moderated discussion among the presenters and between
the presenters and the audience.