Numerous attempts to establish the relative chronology of the songs of the Greek epos, or of particular passages therein, have depended on the linguistic evidence in our received texts. These studies, while producing some suggestive results, have not solved the problem of chronology for early Greek poetry. The language of epic developed in parallel with the colloquial speech of the singers in many ways, but the Kunstsprache has rules and tendencies which work counter to the diachronic development of the singers’ dialect. Also, assumptions about the history of the development of the epics heavily influence the definition of linguistic material as ‘archaic’. As we shall see below, one's model for the development of the songs of the Greek epos, whether through a series of dialectal ‘phases’ or through the diffusion of shared poetic material among various local traditions, has a profound effect on attempts to elicit chronology from the language of the epics.

The mixed dialect of epic

The dialect of the epics mixes archaic forms with innovations; it likewise mixes Ionic forms with forms showing Aeolic phonology and morphology. Linguistic archaisms by definition precede innovations, and concentrations of innovative forms are generally taken to indicate a later work or passage. Still, the presence of any linguistic innovation indicates that a passage could only have been composed after that innovation had occurred. For example, a (metrically guaranteed) ‘neglect’ of digamma could not reasonably have entered epic language until initial [w-] had lost consonantal force in some relevant dialect. Clearly, however, the effects of digamma persisted in epic singing well after the sound itself had disappeared, and thus, any apparent ‘observance’ of digamma, say, need not have entered the text before the loss of the sound in colloquial speech.