Rodda Martina Astrid. A corpus study of formulaic variation and linguistic productivity in early Greek epic

University of Oxford, 2021. — 303 p.This thesis addresses the general issue of formulaic variation in early Greek epic - not formulaic change, in that I do not adopt a diachronic perspective or seek to restore the original form of any formulae, but simply to describe how different variants of the same formulae coexist, and what their patterns of variation can tell us about formulaic language. I introduce the concepts of formula and formulaic variation in part I, which is functionally a two-part introduction (chapters I and II). Part II addresses the issue of syntactic variation, focusing on formulae composed by an adjective and a (common) noun; the behaviour of these formulae is analysed in a corpus made of Homer, Hesiod, and the major Homeric Hymns, and compared to the behaviour of similar structures in a non-formulaic 5th-century baseline corpus. The final chapter in part II (chapter V) analyses the behaviour of the same Adj + N pairs in a Hellenistic poem, the Argonautica of Apollonius of Rhodes. Part III focuses on semantic variation, and applies Distributional Semantics in order to study the range of meaning that objects of transitive verbs can take in Homer and Hesiod. Both Parts II and III have an introductory chapter of their own (chapters III and VI, respectively), in which the linguistic framework that will be used in the whole section is introduced and applied to a smaller-scale study.