Michael Penn: Automated Syriac Script Charts

A digital humanities team of a computer scientist, a dozen research, and myself is developing software that brings together recent advances in digital handwriting analysis with scholarship in Syriac studies. The goal of this Mellon funded project is for a scholar to scan pages from a manuscript of interest. The computer would then compare the test document to those in its database and list the degree of similarity between this document’s handwriting and that of other manuscripts. It would then compare the handwriting with documents of known dates, suggesting a year range during which the manuscript was originally written.

As part of this larger project, our team has assembled the world’s largest collection of securely dated Syriac script. Although the long term goal is to help in writer identification, this database is also extremely useful for the generation of more conventional script charts.

This presentation demonstrates how our system can generate customized script charts. That is, the user can specify variables such as date range, script style, letter morphology, as well as overall chart layout. Using its database of tens of thousands of securely dated Syriac letters, the system then automatically generates the specified script chart. The result is a particular valuable resource for human paleographers trying to estimate a manuscript’s date of composition. Although it would require a new database, with just a few hours of coding the interface can be customized for other languages such as Latin, Greek, and Arabic.