The (Annotated) Language of the Homeric Heroes: Towards a Treebank-Based Approach

The aim of this contribution is to explore whether and how computer-based stylistic analyses, grounded on some of the most traditional techniques in literary computing, can help scholars approach the question of characterization in the Homeric poems. We start by analyzing the distribution of the most frequent words in the direct speeches of 25 characters in the Iliad and Odyssey. The data are taken from the complete morpho-syntactic annotation encoded in the Daphne treebank, a revised version of Perseus Ancient Greek and Latin Dependency Treebank. We investigate the relations and differences among the characters with the help of distance matrixes created with Delta, a popular measure of textual differentiation used in authorship attribution. We briefly explore the material in two different directions: a short series of case studies, involving the distribution of selected frequent discourse particles (ἀλλά and γάρ), and a generalizing approach comparing all major characters in the two poems.