Social Network Analysis of ancient Roman Comedy

• Project currently being developed by:
  • Hans Bork (Assistant Professor of Classics, Stanford University)
  • Marguerite DeMarco (Stanford University)
  • Annie Lamar (PhD Candidate, Classics Dept., Stanford University)
  • Stanford Center for Spatial and Textual Analysis (CESTA, https://cesta.stanford.edu)
  • With previous contributions from:
    • Salma Kamni (Stanford University)
    • Daniel Bush (Postdoctoral Fellow, Stanford Center for Spatial and Textual Analysis)

Project Overview:

This project aims to devise a methodology for generating "social network maps" among characters in Ancient Roman Comedy, using different indices of interaction. For example: total lines spoken between characters, vs. total time onstage vs. total number of characters in a play; and so on.

Plautus is an important figure in Latin literature because (among other things), his work is one of the few places where we can see people of different statuses, ethnicities, economic classes, and genders interacting with one-another. (Most of the Latin literature from ancient Rome tends to feature elite, Roman, male figures; Plautus is a rare exception.) The interactions among Plautine characters are thus important evidence for our understanding of ancient Roman society—but, so far as I know, they have never been examined from a contemporary networking perspective. Until now!

Project Goals

The major short-term goals of the project are: 1) to develop a simple workflow for mapping "social network relationships" in Roman Comedy (initially, the plays of Plautus, with later work on Terence); 2) to make the generated maps open to various queries—i.e., let users query them for various data points, such as "network interactions involving enslaved characters"; 3) to make the generated maps (and associated data) available online, for free, for use by scholars and interested non-specialists.

Long-term goals include opening up the developed framework to ancient drama generally (e.g., Greek and Roman Tragedy; Senecan drama; etc.), and to allow comparison of data among the various mapped data-sets. (E.g., comparing number of interactions by certain character-types across all of Greco-Latin drama.) Advanced use-cases might include e.g. filtering network interactions by defined categories or tags (e.g., "all soliloquies in Roman comedy"; "all instances where non-dramatic entities seem to appear onstage"; and so on).

By working with CESTA I hope to, ultimately, develop a system that will easily make social network maps—and corresponding textual data—of Plautus' character interactions available online, and thus accessible for anyone to use. And, if the system works well, then I could see extending it to other ancient playwrights (of which there are quite a few). This is of course a long-term goal; current work is devoted to developing prototype workflow and methodology, by mapping character relationships in a single play by Plautus ('Captivi'). This text will be a "laboratory" for how we can (and cannot) graphically model character interactions in different ways; the developed methodology will then be extended to all of the other plays in Plautus' corpus.

Example STATIC map:

• Currently we are working to develop a dynamic (i.e., interactive and non-static) network map for the play Captivi, by using the D3.js engine.
• Here is an example of a simple, static network map created using Python, matplotlib, and networkx:

Prototype DYNAMIC map:

• The dynamic prototype is being developed using D3.js and the Jekyll web platform.
• Current Prototype Available Here
  • Repo for the GitHub Pages site available here.

Project TODO:

• Develop a prototype workflow and toolkit to publish SNA data online.
• Develop a project roadmap, based on the existing prototype.
• Develop a plan for applying the prototype to other texts than Captivi.
• Devise a plan for future development:
  • Develop a workflow for contributors. E.g., for contributing to various stages of work, such as data gathering, data encoding, site coding, and so on.
  • Compose and publish 'Project Roadmap' to GitHub page.
  • Recruit individuals to work on the project. E.g., graduate students; colleagues in the field; undergraduates; CESTA interns.
• Advertise the project.
• Investigate feasible publication options based on the project. E.g., perhaps a short article to a DH-specific journal? Or to a more open-minded Classics journal?

Note to contributors

Hmmm...