A multidisciplinary conference was organized at Berkeley in April of 2017, and introduced the goal of building a socio-economic network from the 15,000 Neo-Sumerian texts from Drehem, Iraq, ancient Puzriš-Dagan (2100-2000 B.C.E.). The project brought together archaeologists, cuneiform specialists, experts in text analysis and natural language processing from around the world, country and campus. The workshop delineated a workflow for building a social network database from the digitized text archives, hosted and curated by Dr. Manuel Molina (http://bdtns.filol.csic.es/) and Dr. Niek Veldhuis (http://oracc.museum.upenn.edu/projectlist.html). Despite an unsuccessful NEH application, our ambitious project was able to continue thanks to the support of Digital Humanities at Berkeley in the form of a collaborative research project grant.

We learned from this conference that a name authority is necessary in order to properly disambiguate and deduplicate the named entities of the network. This objective was described on the project page, and was the primary objective of the grant. We are excited to report that the Sumerian name authority is complete for the Drehem texts. And while the results require close reading, this has helped detail a method for replicating the name authority for Sumerian Ur-III texts, using Jupyter Notebooks. We are currently working with this new data together with a URAP team, which has already generated an impressive network database. Adam documented our meetings and preliminary results in a shared Google doc, and shared some of the preliminary results in the Assyriology newsletter, Mar Šiprim.

Project type