It can be asked how one can be so confident in classing Homer himself as an oral singer, for if he differed from Phemius or Demodocus in terms of length, he may also have differed radically in his poetic techniques. The very nature of his verse may provide a substantial part of the answer. The style of the poems is “formulaic”; that is, they rely heavily on the use not only of stock epithets and repeated verses or groups of verses—which can also be found to a much lesser extent in a literate imitator like Virgil—but also on a multitude of fixed phrases that are employed time and time again to express a similar idea in a similar part of the verse. The clearest and simplest instance is the so-called noun-epithet formulas. These constitute a veritable system, in which every major god or hero possesses a variety of epithets from which the choice is made solely according to how much of the verse, and which part of it, the singer desires to use up. Odysseus is called divine Odysseus, many-counseled Odysseus, or much-enduring divine Odysseus simply in accordance with the amount of material to be fitted into the remainder of the hexameter (six-foot) verse. A ship is described as black, hollow, or symmetrical not to distinguish this particular ship from others but solely in relation to the qualities and demands of the rhythmical context. The whole noun-epithet system is both extensive and economical—it covers a great variety of subjects with very little exact reduplication or unnecessary overlap. It would seem that so refined and complex a system could not be the invention of a single poet but must have been gradually evolved in a long-standing tradition that needed both the extension and the economy for functional reasons—that depended on these fixed phrase units because of its oral nature, in which memory, practice, and a kind of improvising replace the deliberate, self-correcting, word-by-word progress of the pen-and-paper composer. Admittedly, the rest of Homer’s vocabulary is not as markedly formulaic as its noun-epithet aspect (or, another popular example, as its expressions for beginning and ending a speech). Many expressions, many portions of sentences are individually invented for the occasion, or at least so it seems. Even so, there is a strongly formulaic and ready-made component in the artificial language that was used by Homer, including its less conspicuous aspects such as the arrangement of particles, conjunctions, and pronouns.

It looks, therefore, as though Homer must have trained as an ordinary aoidos, who began (like most of the present-day Yugoslav guslari) by building up a repertoire of normal-length songs acquired from already established singers. The greatest heroic adventures of the past must already have been prominent in any repertoire, especially the Panhellenic adventures of the Seven Against Thebes, the Argonauts, and the Achaean attack on Troy. Some aspects of the Trojan War might already have been expanded into songs of unusual length, though one that was still manageable on a single occasion. Yet the process was presumably carried much further in the making of the monumental Iliad, consisting of more than 16,000 verses, which would take four or five long evenings, and perhaps more, to perform. This breakthrough into the monumental, which made exceptional and almost unreasonable demands of audiences, presupposes a singer of quite exceptional capacity and reputation—one who could impose the new and admittedly difficult form upon his listeners by the sheer unfamiliar genius of his song. The 8th century bce was in other respects, too, an era of cultural innovation, not least in the direction of monumentality, and huge temples (like the early temple of Hera in Samos) and colossal funerary vases (like the mixing bowls and amphorae in the so-called Geometric style from the Dipylon cemetery in Athens) may have found a literary analogue in the idea of a vast poetical treatment of the Trojan War. But in an important sense Homer was building upon a tendency of all known oral heroic poetry toward elaboration and expansion. The singer does not acquire a song from another singer by simple memorization. He adjusts what he hears to his existing store of phrases, typical scenes, and themes, and he tends to replace what is unfamiliar to him with something he already knows, or to expand it by adding familiar material that it happens to lack. Every singer in a living oral tradition tends to develop what he acquires. There is an element of improvisation, as well as of memory, in his appropriation of fresh material; and judging by the practice of singers studied from the middle of the 19th century onward in Russia, Serbia, Cyprus, and Crete the inclination to adjust, elaborate, and improve comes naturally to all oral poets.

Cumulative poetic structure

Homer must have decided to elaborate his materials not only in quality but also in length and complexity. All oral poetry is cumulative in essence; the verse is built up by adding phrase upon phrase, the individual description by adding verse upon verse. The whole plot of a song consists of the progressive accumulation of minor motifs and major themes, from simple ideas (such as “the hero sets off on a journey” or “addresses his enemy”) through typical scenes (such as assemblies of men or gods) to developed but standardized thematic complexes (such as episodes of recognition or reconciliation). Homer seems to have carried this cumulative tendency into new regions of poetry and narrative; in this as in other respects (for example, in his poetical language) he was applying his own individual vision to the fertile raw material of an extensive and well-known tradition.

The result is much more complex than with an ordinary traditional poem. Understanding the origin and essential qualities of the Iliad or the Odyssey entails trying to sort out not only the separate components of the pre-Homeric tradition but also Homer’s own probable contributions, whether distinguishable by their dependence on the monumental idea or by their apparent novelty vis-à-vis the tradition as a whole or by other means. Dialectal and linguistic components must be identified as far as possible—survivals of the Mycenaean language, for example, or words used exclusively in the Aeolian cities of the west coast of Asia Minor, or Athenian dialect forms introduced into the poems after the time of Homer; so must specific references to armour, clothing, houses, burial customs, political geography, and so on, that are likely to be assignable to the Late Bronze Age, the Early Iron Age, or the period of Homer’s own activity—at the very least to be taken as relatively early or late within the whole range of the poetic tradition down to Homer. These are the tasks of modern Homeric scholarship. Yet such different forms and ideas in Homer are not conveniently separated into distinct sections of the text, which can therefore be assigned to early or late phases of composition. On the contrary, they may coexist in a single (artificial) linguistic form or a single descriptive phrase. Any member of the tradition, not least Homer himself, may, moreover, have chosen to archaize on one occasion, to innovate on another. One result is that the epics are dubious authorities for the assessment of historical events like the attack on Troy or the status of workers, just as they are ambiguous sources for early Greek grammar or theology. Another is that they are not bound to a single worldview or period or mode of perception; rather, they unite judgments and experiences never seen together in “real” life into a whole that is literary but nevertheless revealing of the underlying structure of human existence.