The Voyage of the Slave Ship Sally: Exploring Historical Documents in Context

The Voyage of the Slave Ship Sally is a digital project
developed as part of the work of Brown University's
Steering Committee on Slavery and Justice, which in 2003 was
formed to investigate the university's historical relationship to
slavery and the slave trade. The committee's research included
working with archival materials from the John Carter Brown
Library and the John Hay Library at Brown University. A
significant subset of these materials concerned a single slave
voyage conducted by John and Nicholas Brown in the 1760s,
for which a large amount of documentary evidence survives,
including ship's manifests, invoices and receipts, letters, bills
of lading, and the captain's trade book. The Brown University
Scholarly Technology Group (STG) worked with Professor
James Campbell, chair of the committee, and Patrick Yott,
director of the Center for Digital Initiatives, to develop a
resource through which these documents could be read and
explored.
The goal of the project was to present these materials in a digital
form that would respond to the committee's expressed goal to
"help the campus and the nation reflect on the meaning of this
history in the present, on the complex historical, political, legal,
and moral questions posed by any present-day confrontation
with past injustice." The project thus presented a number of
challenges: the breadth of audience, the need to engage readers
with widely varying levels of familiarity with the historical
background, and the need to offer as many ways as possible of
engaging with and probing these materials, so that they would
yield the maximum possible insight into the social, political,
historical, cultural, and economic circumstances surrounding
the slave trade. The initial version of the site, which was
released concurrently with the committee's report, emphasizes
access to the primary source documents through page images and TEI transcriptions. The second phase of the project's
development, which will be completed by the end of 2007, will
provide more detailed forms of access through specialized
interface tools that draw on additional contextual data being
developed by the project faculty.
This poster and demonstration focuses on three areas of
particular significance arising from this project:
1. The interface design, which emphasized ways of engaging
readers of all skill and education levels and providing them
with contextual information needed to interact meaningfully
with the primary source documents
2. The management of interconnections between TEI
transcriptions, metadata, images, and contextual data, which
support a set of exploratory tools.
3. The exploratory tools themselves, which encourage readers
to probe the historical sources and their context in ways that
go beyond the authoritative information provided by the
site itself.
Interface design
One of the central premises guiding the project's design
was the fact that we were anticipating readers in two
broad categories: those familiar with the historical context who
would be immediately interested in exploring the primary source
documents and able to make some sense of them, and those
coming to the site with no prior knowledge, who might have
no initial interest in the documents at all and no way of
exploring them meaningfully. Readers in the former group
would need powerful tools to allow them to probe the
documents, while those in the latter group would need a more
narrative interface emphasizing the historical context and its
modern significance, and providing readers with a sense of why
the primary source documents constitute an important
intellectual resource. The interface is designed with the aim of
leading all readers, eventually, to the documents via links from
the historical narrative to the relevant documentary material,
but it also provides readers with search and exploration tools
that embody as much contextual information as possible:
glossed timelines and maps, social networks of people, and lists
of commodities involved in the slave trade, all drawn from the
information carried in the primary source documents
themselves.
Data Structures
The documents and contextual information are represented
through a set of interrelated data structures, specifically:
• TEI transcriptions of the primary source documents
• digital images of each document at a variety of resolutions
• METS records representing the structure of each digital
object
• a set of simple XML data structures to represent four
categories of contextual information on persons, places,
events, and commodities named in the texts, to allow for
glosses and information about the interconnections between
items in these groups (such as social networks, family
relationships, classifications of commodities, and so forth).
The TEI transcriptions capture a comparatively large amount
of information about the documents' content, by encoding
all references to persons, places, commodities, and financial
transactions so that these can be indexed, glossed,
regularized, and searched. The transcriptions use this
encoding to reference the contextual information, using
key= attributes which point to the relevant records
describing the person, place, event, or commodity in
question. This allows both for basic hooks to allow (for
instance) links to glosses directly from the text, and also
for more advanced exploratory tools which process this
information and represent it through visualizations and
advanced search mechanisms.
Exploratory tools
The primary source documents for the Sally project are
often individually compelling, but they are most
significant when understood as part of a social and economic
network. The existence of this network, however, can only be
grasped indirectly and fragmentarily through the documents
themselves; the reader needs tools for exploration that allow
patterns and connections to become visible, and that can draw
on the contextual information provided. When completed the
project will include a set of interface tools that facilitate
precisely this kind of exploration. For instance, the reader will
be able to explore the network of people and social relations
that surround the Sally voyage and inhabit its documentary
record--owners, captain, seamen, outfitters, workmen,
employees, enslaved Africans, slave traders. Their documentary
relationships (for instance, the senders and receivers of letters
and invoices) will be visible along with all of the metadata
about those exchanges (place and date of writing, type of
document, etc.) so that readers can construct a detailed and
vivid picture of the vectors of communication through which
this voyage is documented. Similarly detailed information
supports the exploration of the commodities described in these
documents, both those that were explicitly traded for slaves
(rum, cloth, iron bars), and those that were structurally essential
to the trade (weapons, sugar, the ship's provisions).
Visualization tools such as word clouds, manipulable
representations of networks, and generated timelines will give
readers a variety of ways to explore this information in ways that allow an overall grasp of the landscape as well as direct
connections to the documentary evidence at every point.
Conclusions
Because the collection of documents associated with the
Sally voyage is comparatively small (on the order of a
hundred or so), the greatest informational value was to be gained
by an emphasis on detail (through the encoding of the
transcriptions) and on contextual information (through the
associated XML data structures that represent ontological
relationships between the various persons, commodities, and
so forth). This kind of approach is uncommon for collections
of historical documents, which characteristically emphasize
scale (consider for instance the Valley of the Shadow, or the
Making of America projects). Using visualization tools to model
and interact with information of this type is also comparatively
rare; visualization tools are being increasingly used for
exploration of very large data sets (often not encoded in XML
at all) but are less common for small, heavily tagged document
collections. This project thus offers opportunities for an
exploration of digital methods which may yield some useful
results at its conclusion--in particular, on the question of how
much value is added by detailed markup and tools that exploit
it.