Mapping Pliny's Social Network: A Case Study in Digital Prosopography

Benjamin Hicks

bhicks@princeton.edu

Princeton University, United States of America

Introduction

This proposed short paper examines the progress and some preliminary
observations of a practical attempt to apply digital humanities methods to the
letters of a second century Roman aristocrat Pliny the Younger. The preliminary
results of this study are

available at the Pliny Project site. Its aim is to present

a case study of how an initial research idea can be expanded to connect with
larger digital humanities work in a particular field.

These letters, written in the early second century CE, are a treasure trove of
social and literary information about the Roman elite during the period
in which Roman territorial control reached its apex. As one of the most
extensive collections of letters from the ancient world, as well as one of the
most thoroughly explored, they are a rich data set on which to draw. They are
paralleled by only a handful of similar letter sets from the Roman world, such
as the letters of Marcus Tulilus Cicero or the four century orator Liba-nius.

In a broader context, Classics has been a frequent and early adopter of digital
humanities methods (see Online Coins of the Roman Empire, the literary
comparative tool Tesserae, and the venerable Perseus Digital Library). The
field has also produced some initial attempts to bridge various “people
indices” into a standard prosopography (i.e. the Standards for
Networking Ancient Prosopographies, hereafter SNAP). Pliny’s correspondents
have been integrated into such resources, but they are either often limited to
major university research collections or so unwieldy as to make consultation
difficult. Examples of this tension include the standard tool for the
prosopography of the Roman Empire, the Prosopographia Imperii Romani, 2nd
edition (de Gruyter: 1933-2015) (= PIR^2), which has reached a massive eight
parts with numerous fascicles. Only the index is widely available online.
Likewise the most recent work on Pliny's names, Anthony Birley's Onomasticon to
the Younger Pliny (Birley, 2012), exists as a traditional monograph, albeit
with a searchable PDF.

Such resources, though tremendously important to scholars specializing in the
field, often constrain access for the broader academic community. Moreover, the
rich information they provide is not structured in a way for easy search and
access.

The research project on which this paper draws developed from a November 2015 -
January 2016 affiliated fellowship at the American Academy in Rome, which gave
me access to the prosopographic material in a single, well-organized location.
Its primary objective was, and is, to create a comprehensive resource for
Pliny's social network with an emphasis on the social class of his
correspondents. My initial inquiries centered on compiling a list of Pliny's
correspondents and attempting to identify them as best possible.
The conventions of Roman naming, which resulted in many similar names within
family groups, renders this difficult. The use of only a single name (compared
to the somewhat standard use of two names) in one of the surviving manuscripts
of the letters further complicates the task. Even if a family and identity of
an individual is known, his or her social standing may not be clear. The Roman
distinctions between a common citizen, the middling administrators of the
equestrian class, and the upper rungs of the senatorial class were very sharp
to them—so sharp they often saw no need to clarify who was of what class for
posterity.

This made data modeling and cleaning a significant challenge, for which I
employed exploratory tools such as SocNetV and more recently Cytoscape, for
exploratory visualization. (Note: this issue of authorial ambiguity is not new
to DH and letters. It has been frequently confronted by projects such as
Stanford's Mapping the Republic of Letters). Some preliminary results are
available through Pliny Project (see above), but they have been revealing both
in terms of confirming known associations and providing new clarity into the
possible editorial methods Pliny used in selecting his letters, which were
curated for publication within his lifetime.

In order to construct a data set for such an analysis, I attempted to model a
degree of closeness of connection by assigning a weight based on the number
of times Pliny either mentioned someone in a letter or wrote to them to a
reciprocal connection. This was saved in GraphML format to construct a diagram
of centrality with shading of points to indicate the social class of Pliny's
correspondents.

My talk will, in addition to discussing the above methodology in greater
detail, center on two examples of preliminary results from this research, and
then turn to future plans for the project. First, the set of social
acquaintances that have often been associated with as what is informally called
“Pliny Country”— near his home near modern-day Como, Italy—and the set
associated with the city of Tifernum, both appear clearly in the social network
map as a set of closer intimates, largely from the same equestrian class
of which Pliny's family originated (the original formulation comes from the
work of the eminent historian Ronald Syme, see Syme, 1991, for his collected
works and the exploration of some of Pliny's connections of Tifernum in
Champlin, 2001).

This gives preliminary confirmation that the methodology of simple weighting
based on mentions as some approximation of closeness can be used in analyzing
his social network.

A secondary observation is a series of correspondents to whom Pliny writes
roughly two to three letters in the second to outer circle of his
acquaintances. Some of these individuals are men who had held the consulship,
the highest office to which a Roman not of the imperial family could aspire and
all were of the senatorial class. Pliny rose to that same class from middling
origins during his career, thanks to the patronage of his uncle and adoptive
father. That there is a cluster of these letters with a remarkably similar
number speaks to an editorial hand at work in their selection. While at this
point identifying a motivation is primarily speculative, at the least we can
say that it reveals a trend not previously identified and demonstrates an
editorial concern for cultivating Pliny's prestige by association.

In addition to the specific application of this data to my own research, the
longer term goals of this project are to provide this same dataset, edited and
curated, to the broader scholarly community. I have currently published a
simple database interface that allows users to search for Pliny's
correspondents and note which letters are written to them. While this may
seem on the surface a straightforward question, by integrating current
scholarship and attempting to identify correspondents fully, it presents new
and easier access for scholars, regardless of institutional affiliation.

Nevertheless, the initial search functions, which allow a name search and a
tentative search by social class, are not sufficient to realize the goals of
the project. My current development work is focused on transitioning the
database to using Django's web application functionality and object database
modeling to allow for the relationships noted in my social network analysis to
be available and searchable. This transition to a standard platform will also
lead to a web application that can be cloned from a tool such as GitHub
and used by DH scholars to build or innovate using my dataset. It also
acknowledges the need to connect this new structuring of the prosopographic
corpus for Pliny to the broader initiatives to create people indices by
including links to PIR^2 search masks and SNAP.

Such an approach takes the traditional field of Pli-nian prosopography and
attempts to open it to a wider scholarly audience. It also emphasizes the
importance of exploratory visualization techniques in examining datasets for
novel connections. This discussion will offer the audience an opportunity to
consider how a project focusing on a specific area of research can connect with
larger scholarly endeavors in DH.