The Homeric poems are often viewed as oral epics that have been transposed into writing. Commentators have dwelt on the presence of certain features, such as the formulaic expressions (epithets and repeated verses) that they see as typical of oral forms. Yet, while the regular repetition of phrases is found in such forms, some have argued that the precise format of Homer’s formula—as defined by the American scholar Milman Parry, who in the 1930s recorded epics in Yugoslavia performed by guslari—is very likely an early literate device.

The epic itself is a case in point. It is often assumed to be a typical product of oral cultures, being sung by bards at courts or in camps. However, records of epics in purely oral cultures are sparse. Epics tend to be found in early states with important warrior classes that enjoy hearing of the brave exploits of their predecessors. These societies already have writing, but the texts are often committed to memory and reproduced by the speaker rather than read aloud from a written version of the text. These oral texts are recited in gatherings of chiefs and warriors by specialist bards and, in fact, are the works that have been written down at some point. Indeed, only because they have been written down are they known at all. And it is a regular characteristic of early written works, such as Homer’s or the Vedas, that, although written versions exist, these are learned by heart, internalized, and reproduced through the spoken word alone, just as is often the case with religious texts such as the Qurʾān and the Bible. At the very moment in history when writing allows one to dispose of verbal memory as a means of recalling such works, the role of verbal memory is in fact enhanced—hence part of the difficulty in deciding whether these works are both orally composed and orally reproduced. They are typical of societies where only the few can read and where mnemonic skills and devices that encourage the perfect oral reproduction of written texts are therefore encouraged.

The way that purely oral forms have been changed under the influence of written cultures is illustrated by the Kalevala, the Finnish national epic, a final version of which was published in 1849 by Elias Lönnrot. A systematic collector of folk poetry, Lönnrot concluded that what had been recorded as distinct poems could be conflated into a continuous folk epic. He joined a number of shorter compositions together with material of his own and imposed on the whole a unifying plot. Its publication had an enormous influence on Finnish culture. It has been suggested that the Gilgamesh epic may have been constructed in a similar way; other cases have been reported from Africa. In North America stories that centre on particular characters such as Coyote or Raven (see Raven cycle) are sometimes grouped together by scholars into cycles, but it seems doubtful whether these cycles represent a meaningful category for the Native Americans who created these tales.

Differences between oral and written literatures

Oral and written literatures differ in their authorship and audience. In oral cultures the memory of authorship, though never entirely absent, is of little general importance—occasionally with songs but not with myths, folktales, and, rarely, epics. That is not to say that these genres do not become the subjects of intellectual property rights. Songs may be associated with particular clans, recitations with specific groups or gatherings. But usually no individual author is traced. That absence, however, does not imply that there is a process of collective composition. Each reciter will introduce variations of his own, some of which will be taken up by succeeding speakers for whom the previous version will have been the (or a) model. In this way changes are constantly being introduced by individuals but anonymously, in a syntagmatic (contextual) chain, without looking back to any fixed original. Only with writing and with oral recitations in literate societies, as in the case of Homer or the Vedas, is the oral transmission of an original possible, partly because writing introduces a new dimension to verbal memory and partly because reference can then always be made back to the “correct” version. As a result, it seems to be in these early literate societies that the development of mnemonic skills and aide-mémoire is first found.