Thinking Archivally: Search and Metadata as Building Blocks for a New Digital Historiography

1
Thinking Archivally: Search
and Metadata as Building
Blocks for a New Digital
Historiography
Sternfeld, Joshua
jsternfeld@neh.gov
National Endowment for the Humanities, USA
"Records are no longer fixed, but dynamic. The
record is no longer a passive object, a 'record'
of evidence, but an active agent playing an on-
going role in lives of individuals, organizations,
and society."
1
Advances in digital representation and
preservation have ushered in new perspectives
for defining the record. Terry Cook, speaking
on behalf of a growing cohort interested
in reshaping the disciplinary boundaries of
archival studies, argues that records no longer
possess the aura of absolute authority that
they once held. Practitioners and theorists
working within a postmodernist framework
have broken open the once sacred bond between
the historical craft and the archive, in the
process challenging notions of evidence, truth,
and narrative.
One of the great beneficiaries and active
participants of this re-evaluation of the
archive and the record, of course, is digital
history. Digital tools, from Omeka to ArcGIS,
have empowered a growing community of
professional and amateur historians, museums,
and libraries to provide unprecedented access to
collections of primary materials and historical
data. Such tools also demonstrate the ease with
which archival and historical practices have
come into contact with one another, thereby
disrupting conventional understanding of the
record in both fields.
In this paper, I will address the cross-
disciplinary relationship between history and
archival theory as one component in the
broader development of a much-needed digital
historiography.
2
I will argue that principles
of archival theory and historiography together
may guide the evaluation of digital and new
media historical representations, especially with
regards to the contextualization of historical
evidence. Whether considering an online
archive, database, or GIS visualization, the
aggregation of large data sets necessitates
proper archival management of the data.
Besides enhancing the long-term sustainability
and preservation of the representation –
itself a worthy and often overlooked objective
–, the application of archival standards to
data collection, organization, and presentation
influences the type and quality of conclusions
users may generate. Within this complex
association among history, archival theory,
and digital technology this paper will examine
two interrelated "building blocks" – search
and metadata – that work hand-in-hand to
form the foundation for a sound digital
historical representation. How a user queries
a representation, even one that is non-textual,
governs the quality of historical knowledge at
the user's disposal, whereas a representation's
content metadata governs the conclusions the
user may draw with that knowledge.
1. A Call for Digital Historiography
New techniques to query, sort, catalogue, and
visualize historical data have brought renewed
interest to understanding the past on every
scale, from the personal to the global. As
scholars and teachers, we have encouraged
digital exploration, whether in gathering local
data with the support of a historical society
or archive, or repurposing historical materials
through museum installations, websites,
documentaries, and multimedia mashups.
Despite promising possibilities in historical
computing, the emergence of digital history
has also created a distinct fissure in the
wider field of history. Practicing digital history
challenges methodological preconceptions.
Conducting search queries across vast digital
collections seems antithetical to visiting
an archive. Similarly, navigating through
a three-dimensional environment enhances
interactivity and engagement with the historical
representation, and in the process confronts, or
at times abandons altogether, the core activities
of reading and writing historical texts. In short,
history in the digital age has upended notions
of representation, context, inquiry, narrative,
linearity, temporal and spatial orientation, and
experience.
3

2
This undeniable shift in the landscape demands
that we harness the potential of digital history
while not altogether abandoning established
theoretical and methodological practices. A
rush to embrace new digital modes of doing
history, unfortunately, has overwhelmed a
parallel critical examination of changes to
these fundamentals. The same techniques and
technologies that are laudably tearing down
institutional barriers, challenging entrenched
theories, and introducing new voices and
democratic perspectives, can also advance
specious information and theories; distort or
obscure the historical record; or worse –
eliminate it altogether. The role of the historian,
therefore, has shifted from that of exclusive
authority to the equally critical role of mediator
of historical knowledge. If active participation
and exploration have become the benchmarks
of digital historical representations, then the
(digital) historian must ensure that the manner
of user participation is conducted equitably and
responsibly insofar as the knowledge produced
through the representation is predicated upon
rigorous logic and concentrated historical data.
What principles should a new digital
historiography advocate and why is its
cultivation imperative? A working digital
historiography will enable critical engagement
with digital and new media representations, a
challenging endeavor considering the spectrum
of possible forms that a representation may
take. We may justifiably question whether an
online collection, for example, shares traits
with a GIS-based visualization. While each
representational genre warrants a unique set
of evaluative criteria, commonalities across
formats and historical content do exist and
warrant further attention. We may begin with
the notion that all representations possess some
form of a user interface. Interrogating the user
interface can lead one to assess the transparency
with which the representation has selected
and organized its content. We may also ask
whether its formal design complements and
provides sufficient access to the content. With
a scholarly text, answers to such questions
are readily apparent by poring over indicators
such as footnotes, bibliography, and the table
of contents. Many digital representations,
however, collate information within multi-
dimensional, non-linear structures, thereby
subverting or eliminating such identifiable cues.
As Edward Ayers remarks, "We cannot judge a
Web site by its cover – or its heft, its publisher's
imprint, or the blurbs it wears."
4
2. The Building Blocks of Digital
Historiography: Search and
Metadata
In developing a set of evaluative criteria,
we must consider the association between
a representation's form and content, which
together comprise the representation's overall
historical argument. While there are numerous
components worthy of consideration, two in
particular – search and metadata – determine
to a large extent how a representation
organizes its historical information. Without
a robust search engine the user cannot
access historical data; similarly, without quality
metadata, a strong search engine is rendered
ineffective. While this may seem self-evident,
the integration of search and metadata in a
representation runs much deeper; it affects,
and is affected by, nearly every aspect of
the representation, including its interface,
aesthetic, design, structure, and functionality.
Search and metadata together govern the
transformative process by which historical
information becomes historical evidence.
This paper will use examples of current digital
collections and visualizations to illustrate how
search and metadata contributes to the overall
value of the representation. I will argue that an
assessment of these two building blocks, when
considered from both an historical and archival
perspective, can shed light on the argument
put forth by the representation. In the case of
an online collection, for example, the creator
must weigh the benefits of generating metadata
according to standardized thesauri, scholarly
input, or folksonomy. These very different
approaches, if applied to the same archival
collection, would not only influence the type of
audience that may use the archive, but also steer
users towards divergent search results, which
could ultimately determine how the content is
recombined.
5
A reconstruction of an historic building,
meanwhile, invites a "search" process of a
different sort. Searching occurs while the user
navigates through the environment. Is the user
invited to discover new sightlines or gauge