Digital Initiative for Classics: Epic Speeches (DICES)

About DICES

DICES was founded in 2018 by Simone Finkmann, Christopher Forstall and Berenice Verhelst as a collaborative international research project centered around the creation of a comprehensive database of direct speech in Greek and Latin epic and the exploration of new avenues for research on speech in the epic tradition, from Homer to late antiquity. The project seeks to inspire future research in the field, promoting Digital Humanities methods, drawing on theoretical insights from the fields of social psychology, the study of emotions, and narratology, and expanding the epic canon.

The DICES database, launched publicly in December 2025, can be accessed in its most up to date version via this webpage. It is designed to be dynamic. New data and corrections will be added over the years.

The database is designed to be consulted, on the one hand, via its user-friendly web interface , for browsing, searching and filtering through our structured metadata, currently covering about 5000 speeches and more than 1000 characters in 52 Greek and Latin poems. On the other hand, our database design equally facilitates machine-driven search and retrieval of speech and character records using the general-purpose computer programming language Python. Example code and tutorials will be added to the digital appendix section of this site.

Alongside the database, DICES presents a collection of case studies, innovative research on direct speech in Greek and Latin epic, by an international group of scholars who acted as the database's test users. The resulting peer reviewed edited volume is published open access by Brill (Forstall-Verhelst 2026): Direct Speech in Greek and Latin Epic Expanding the Methods and Canon.

This site also hosts the digital appendix to the volume: supplemental materials provided by individual chapter authors, such as raw data, summary results, and computer code. It includes interactive demonstrations, tables, and visualizations created for several chapters whose results could only be represented in a limited form within the printed volume.

An archival snapshot of the database has been deposited with Mount Allison University Libraries and Archives and is accessible through Borealis, the Canadian Dataverse Repository, at https://doi.org/10.5683/SP3/N8LS2Y.

Mt. Allison University University of Amsterdam

Objectives

Project Team

This work is supported by