MATMIDAH (Machine Analysis of Talmudic Memrot, Inference, Dialogue, Argumentation, and Hermeneutics)

• In Event: Jewish Studies and Digital Humanities Workshop

Abstract

This paper presents MATMIDAH (Machine Analysis of Talmudic Memrot, Inference, Dialogue, Argumentation, and Hermeneutics), a computer program that extracts Tannaitic and Amoraic quotes from punctuated text of the Talmud. MATMIDAH creates a relational database of these quotes together with associated information, such as the quote location (tractate and daf), author (including the chain of transmission, if present), the phrase used to introduce the quote (אמר, בעי, דרש etc.), to whom the quote was directed if this information is present, and the punctuation used to terminate the quote (indicating whether the quote is a question or statement). Use of a relational database as a repository for the extracted information allows ad hoc queries to be answered instantaneously. For example, a researcher can get an immediate answer to the question of which pair of Rabbis constitutes the most frequent addresser, addressee structure in the Talmud. (Spoiler alert: רבינא-לרב אשי . On the other hand רב אשי only addresses רבינא a couple of times!) Together with a companion database of Rabbis which includes their names, generations, and place of residence, this database can be used to address a number of subjects statistically, including, for example, social network analysis among the Rabbis, distribution of activity by generation, tractate, and type of activity (i.e. question vs. answer vs. independent statement), and philological comparisons between individual Rabbis and generations. The paper describes first results obtained in these areas. Additionally, the paper discusses the methodology used to construct the database and difficulties attributable to unique Talmudic structures and source materials. Finally, the paper proposes directions for further work.
While this project is presently under development, I anticipate that an interface to the database will be openly accessible on a website, once the project has reached a first stage of completion.

Author

• Michael Sperling, New York University