Archaisms and Innovations in the Songs of Homer

The dissertation comprises three case studies on the history and prehistory of Homeric language, focusing on the ways in which archaic forms are preserved, and innovated forms created. In the first study I examine Homeric accentuation, together with related issues of morphology and morphophonology, in the u-stem adjectives. Beginning from the archaic oxytones 'close-set' and 'thick', I outline the historical developments leading to the paradigmatic feminines in - , which are based on the masculine/neuter stems, and to the recessively accented adjectives 'abundant', 'wooded', 'sweetly sonorous'. I propose that the recessive accent results from the morphological isolation of these words (i.e. they lack a masculine/neuter base), coupled with a subsequent re-accentuation to the default, recessive accent of the language. Turning to Vedic, I will examine its cognate class of adjectives, whose accent is unequivocally oxytone; for instance sva d-i v- 'sweet' is the masculine/neuter stem to sva d-v-i , the feminine. But the morphophonology of the u-stem adjectives requires further study, I argue, and must be set in the broader context of Vedic accentuation. Returning to Greek, I look into a few nouns arguably going back to substantivized adjectives, arguably reflecting zero-grade ablaut of the suffix. Such nouns would correspond precisely with Vedic, where zero-grade ablaut of the suffix is the rule (Ved. -vi ) : 'fathom, span of the arms', 'street', and possibly a few others. Taken together, these accentual classes chronicle the history of u-stem morphophonology in Greek. In the next case study I treat how innovations and archaisms developed within one morphological category, the compound s-stem adjectives. In particular, I investigate anew questions of accents and of ablaut grades: which are archaisms, which innovations? To do so, I offer a revised philological account concerning the various accentual classes of s-stem adjectives, then argue that the recessively s-stem adjectives agree most closely with the largely overlooked Indo-Iranian evidence. Re-examining the evidence for Greek accentuation offers in turn an opportunity to look again at the evidence for archaisms and innovations in Greek ablaut. Greek evidence from zero-grade ablaut in the root of second compound members, such as 'terribly suffering', sometimes understood to reflect ancient PIE derivational processes, reflects rather a highly significant innovation in Greek morphology: the class of sstem adjectives transforms from a denominal to a deverbal class. I will demonstrate that the zero-grade ablaut in the second member reflects the verbal bases from which the adjective derives (in this case the aorist 'to experience; suffer'). Why the aorist, opposed to the present or perfect stem, so often serves as the verbal basis in deverbal derivation will be a question I can pose, but cannot fully answer. Finally, I will work through the Indo-Iranian- effectively just Vedic- evidence for accent and ablaut in the cognate class of s-stem adjectives. I will establish first a philologically sound position for the varying accentual classes in Vedic, then will ask in what ways the Indo-Iranian evidence corresponds to the Greek. This re-examination of the combined evidence of Greek and of Vedic leads to a substantially revised picture of the derivational morphology of s-stem adjectives in the protolanguage. The last study casts a wider net, turning to issues in the transmission of Homeric poetry across Greek dialects and across generic boundaries. I focus the case study on one form found in one formula, 'in mind' in the hemistich , incontestably the older form of the dative plural of (for Cl.Gk. ), but only contestably "Homeric". The hemistich with is inscribed on a funerary monument in Attica, but paradoxically may not be evidence for the Attic dialect at all: with a-vocalism closes a Homeric verse-end formula (Hom. ), but in Homer only is ever found; and is unknown to all other Attic documents, while found abundantly- and more abundantly than the lexica and handbooks let on- in texts of the Doric West (Pindar, Stesichorus, and the Orphic leaves). In our study, complications of language and genre come to the fore: Why use a Doric form in an Attic epigram? Why use a Homeric formula in an elegiac couplet inscribed upon a funerary monument?