Time Series Modelling in the Analysis of Homeric Verse

Abstract

Among other techniques of oral composition, the rhythmicity of hexameter seems to play a role in memorising and re-composing in performance long sections of Greek epic poetry. The article attempts to explain what feature of hexameter makes it rhythmical and thus suitable as an epic meter. Homeric verse was analysed using the ARIMA method. Roughly a hundred ten-verse samples from the Iliad were coded as binary sequences of two types. In one type of coding, verses were represented as sequences of long and short syllables, while in the other type, assuming the existence of ictus (metrical stress), as series of dynamically stressed and unstressed ones. The results, obtained using estimated linear models of stochastic processes, clearly suggest that the stress-based series are more rhythmical than the quantity-based series. This can be taken as an argument in favour of the reality of ictus in Greek hexameter.