c3-fig-0001 — Map 3.1 Greece and Asia Minor.

CHAPTER 3
Orality and Literacy

Ancient Greek Literature as Oral Literature

Steve Reece

For most of its history humankind has been illiterate. But this does not mean that humans used to be literarily less sophisticated: they sang hymns to gods, chanted ritual curses, ululated funerary laments, crafted complex genealogies, invented proverbs, fables, and folk-tales, and even composed long heroic epics in demanding metrical forms. All these genres of literature were orally created, orally performed, and orally transmitted to subsequent generations. The inhabitants of the lands that we now know as Greece were no exception. All ancient Greek literature was to some degree oral in nature, and the earliest literature was completely so. Consequently the term “ancient Greek literature” in a literal sense of the words is an oxymoron. For “literature” requires litterae “letters”: letters written by an author, whether with a reed pen, a printing press, or a keyboard; letters read by a reader, whether on a roll of papyrus, a sheet of paper, or a computer monitor. And these are activities foreign in greater and lesser degrees to the composition, reception, and transmission of much of what we by convention call ancient Greek literature.

A chapter on the orality of ancient Greek literature should properly entail a survey of almost all of ancient Greek literature, and it would require one to proceed accordingly, beginning with Greek epic, which was orally composed, orally performed, and orally transmitted for many generations before anyone felt the need to write it down. We know this for many reasons, most obvious of which is the simple historical reality that during the early stages of Greek epic, between the twelfth and eighth centuries BCE, Greece possessed no writing system. The Cretan syllabic writing system that Greece had used for some centuries before this time was adequate for record-keeping but was not suitable for the recording of Greek epic, even if, as is unlikely, it had crossed anyone’s mind to do so; and Greece had not yet adapted the Phoenician writing system that it first used to record various forms of poetry, including epic, and that it still uses today. Greek lyric, both choral and monodic, which was sung to the musical accompaniment of a lyre, had a similar history: the surviving remnants of Greek lyric are from a period well after the first attestation of Greek epic, but lyric reveals, by the great antiquity of its meters, such as the glyconic and pherecratean, that as a genre it actually predates the epic hexameter. Thus, it is no surprise that the Homeric epics themselves mention various types of songs associated with choral lyric: marriage hymns, funeral dirges, harvest songs, the paean, etc. The rise and early stages of lyric, then, like epic, can be placed well back into the pre-literate period of Greek history. Greek iambic verse too must have flourished long before its first attestation in writing in the seventh century BCE. Iambic verse was used for ritualized insult and obscenity, activities not prominent in the more heroic epic verse form, yet surely contemporary with it (we get a glimpse of the genre in the episode of Thersites in Iliad 2). Greek elegy, though later associated primarily with the silent epigrams inscribed on gravestones, was in its earlier form orally performed to the musical accompaniment of a pipe. Greek drama, too, was fundamentally oral: both tragedies and comedies were composed to be performed orally, with spoken, chanted, and sung parts, and they were intended to be received visually and aurally, by an audience at a particular festival on a particular date. Although scripts were composed before the actual performances, actors were free to manipulate and interpolate, and although these scripts were copied and eventually made their way into collections of texts, and even textbooks, in origin the dramas were intended to be performed but once – to a live audience (cf. Baumbach, this volume, ch. 22, p. 344–52). We could go on: Greek oratory entailed speech before a live audience that was timed by a water clock; Greek philosophy was often presented in the form of an oral dialogue; even the quintessential literary genre of Greek history was designed to be read aloud to an audience of listeners – Herodotus is said to have made a fortune by offering public readings of his work in Athens.

In the late Classical and early Hellenistic periods we can begin to speak more properly of writers and readers of literature (cf. Nünlist, this volume, ch. 19, pp. 296–301), but even then it was normal to read aloud, sometimes privately but usually publicly. Literacy rates remained low throughout Greek history, at least by modern standards, and books lacked the aids to reading that we today take for granted – upper- and lower-case letters, word division, punctuation, sentence and paragraph demarcation, line and page numbers – so the reading of literary works remained the domain of highly trained specialists, and the reception of these works for most people was an aural and communal event. The solitary, silent reading with which we are so familiar today would have struck most ancient Greeks as peculiar, at least in the early period, and someone who conducted himself in such a manner would have been regarded by many as an ἰδιώτης (idiōtēs “private person”).

Classicists have not always fully appreciated this oral dimension of ancient Greek literature. Being philologists and bibliophiles, we have traditionally regarded textual criticism as the centerpiece of our discipline; we have been eye philologists rather than ear philologists. More attention to various aspects of historical performance, such as cultural setting and audience reception, has changed the discipline as a whole, and this is nowhere more true than in the study of ancient Greek epic. Our focus in this chapter, then, will be on the ancient epic verse form, not just because it is the earliest attested Greek literature, but also because it illustrates so paradigmatically the fundamentally oral dimension of Greek literature. Moreover, a survey of the history of Greek epic is a case study in how a poem orally performed and aurally received during the earliest period develops over time into an exemplar of a written text, to be dissected and analyzed by teachers and students as a staple of study in the Hellenistic school systems. In short, a survey of the history of Greek epic well illustrates the powerful dynamic between orality and literacy that was at play in all genres of literature throughout Greek history.

1. Oral Features of Ancient Greek Epic

A century ago it was not uncommon to approach the Homeric epics in much the same manner in which one would approach other great literary epics: Vergil’s Aeneid, Dante’s Divine Comedy, Milton’s Paradise Lost, and even more modern epics such as Tolstoy’s War and Peace. Critics commonly isolated exemplum and imitatio, identified textual allusions and verbal irony, and praised Homer’s originality and selection of les mots justes, as though the individual genius of the author was imprinted on each page of text. Yet, as some realized, even from the earliest period, there was something very different about Homeric epic.

The language (cf. Willi, this volume, ch. 29, pp. 445–7) was very repetitive: ornamental epithets are attached to all the prominent characters (“swift-footed Achilles” occurs 33 times in the Iliad; “much-suffering Odysseus” occurs 37 times in the Odyssey); half- and whole-line formulae describe the most common actions (“So he spoke, and all of them were stricken with silence” occurs ten times in the Iliad, five times in the Odyssey; “They put their hands to the good things that lay ready before them” occurs three times in the Iliad, 11 times in the Odyssey); entire speeches are repeated almost verbatim (Agamemnon’s promise of rewards to Achilles in Iliad 9.122–57 and again in 9.264–99). Not just words but entire scenes were very stereotypical in nature, with close verbal and structural similarities, especially scenes that narrated frequently occurring activities in the epics: arming for battle (Il. 3.328–38, 11.15–46, 16.130–44, 19.364–91); preparation of feasts (Od. 1.136–40, 4.52–6; 7.172–6, [10.368–72], 15.135–9, 17.91–5); as well as sacrifice, libation, dressing, bathing, bed-preparation, decision-making, and so forth. Episodes tended to occur as doublets and triplets: e.g. Odysseus’ two weeping scenes in Scheria; the suitors’ three peltings of Odysseus in his own palace. Likewise, character doublets were very common: e.g., Circe–Calypso; Eurycleia–Eurynome; Melanthius–Melantho.

Similarities in details and sequences of episodes also occurred in larger themes and narrative patterns, such as the combat scenes, in which the following underlying structure can be detected: 1) comrades urge each other on; 2) horses are yoked to chariots; 3) arms and armor are catalogued; 4) the advancing troops are described by a simile; 5) prayers and libations are offered to the gods; 6) genealogies of a pair of combatants are revealed; 7) taunts and counter-taunts are shared by combatants; 8) an individual combat is described, in which warrior A throws a spear at warrior B but misses; B strikes A with his spear, but it glances off; A kills B; 9) B, the defeated warrior, “bites the dust,” and his soul goes flying off to Hades; 10) A, the victorious warrior, vaunts over the body of his victim and strips his armor.

Finally, ring composition and other framing devices held the epic narrative together, both on the smaller level of the individual scene and on the larger level of the monumental epic. Simple ring composition involving just a few verses can be observed, for example, at Iliad 2.688–94 and 6.269–79. An example of extended ring composition can be observed in Diomedes’ speech to Glaucus at Il. 6.123–43: A. What mortal man are you? B. Mortals who face me perish. C. But if you are a god … D. I would not fight gods. E. Lycurgus did not live for long after angering the gods. F. [story of Lycurgus]. E. He did not live for long, since he was an enemy to the gods. D. I would not fight gods. C and A. But if you are a mortal … B. Come close and perish. Framing devices hold together larger segments, and even the entire epics: the various adventures of Odysseus’ apologoi are arranged concentrically around the Hades episode; the entire Iliad is symmetrically arranged, with the themes and episodes of Book 24 balancing those of Book 1. The story of the realization of the true causes behind these peculiar features of Homeric style is an interesting and inspiring one, so it is worth telling in some detail.

2. Oral-Formulaic Theory and Homeric Epic

Homeric studies in the late nineteenth and early twentieth centuries was taken up largely with the “Homeric Question”: the who, what, when, where, why, and how of the origin of the Homeric texts. The Unitarians naively effaced many real problems that resided in the Homeric texts in their romantic conviction that there was one poet who created these epics ex nihilo. The Analysts vigorously divided the epics up into various lays by different poets plus an array of later interpolations (Poet A, B, C, R, etc.), but they could not agree even among themselves: some offered the view that Homer was the poet at the end of the process who compiled a group of independent earlier lays rather incompetently into one large poem, contributing only the “glue” that kept these individual lays together; others preferred the view that Homer was the poet at the beginning of the process, the brilliant creator of a very ancient short epic to which considerable additions were later made.

Into this quagmire strode Milman Parry, a 21-year old graduate student at the University of California, Berkeley, whose work would eventually render the controversy between Unitarians and Analysts irrelevant. Both were right, and both were wrong. As the Unitarians claimed, a single poet did compose the entire epic, possibly even both epics. But, as the Analysts had realized, this poet was indebted to many earlier poets for his tales and even for the language in which he narrated them. Homer relied upon a long tradition of oral poetry that provided for him the very words (epithets, formulae, type-scenes) that were the building blocks of his epics. First in his master’s thesis at Berkeley, then in his doctoral dissertations at the Sorbonne, written under the guidance of the French metrician and linguist Antoine Meillet, Parry proved beyond a shadow of a doubt that Homer’s diction was traditional and inherited. This proof was based on his discovery that the system of epithets attached to almost every character and object in the epics was characterized by the features of what he called complexity (also length and extension) and economy (also thrift and simplicity) [the terms most often used today are scope and economy]. By complexity Parry meant that each prominent character and object in Homer had been endowed with an array of epithets so as to be able to fit into any common metrical space: e.g., bucolic diaeresis to line-end; fourth foot male caesura to line-end; third foot male or female caesura to line-end; line-beginning to third foot male or female caesura; full line. Thus every major character could be accommodated effortlessly into a variety of metrical situations. Consider, for example, the common epithets of the hero Hector when in the nominative (subjective) case:

– ˘˘/ – ˘˘/ – ˘˘/ – ˘˘/ φαίδιμος ‛Έκτωρ (phaídimos Héktōr)
– ˘˘/ – ˘˘/ – ˘˘/ – κορυθαίολος ‛Έκτωρ (koruthaíolos Héktōr)
– ˘˘/ – ˘˘/ – ˘ μέγας κορυθαίολος ‛Έκτωρ (mégas koruthaíolos Héktōr)
‛Έκτωρ Πριαμίδης (Héktōr Priamídēs) ˘˘/ – ˘˘/ – ˘˘/ – ×
‛Έκτωρ Πριαμίδης βροτολοιγω˜̩̩ ı̓˜σος ̓Άρηϊ (Héktōr Priamídēs brotoloigôi îsos Árēï)

In Parry’s view, the choice of epithets – whether Hector is to be described as “illustrious,” “shining-helmed,” “great shining-helmed,” “son of Priam,” or “son of Priam equal to Ares destroyer-of-men” – was determined not so much by the context of the passage as by the requirements of meter.

By economy Parry meant that generally only one epithet for a character or object was available to fill each common metrical space: thus Hector is never referred to as φέρτατος ‛Έκτωρ (phértatos Héktōr “pre-eminent Hector”), for example, because the metrically equivalent φαίδιμος ‛Έκτωρ (phaídimos Héktōr “illustrious Hector”) is adequate for the poet’s purposes (the metrically equivalent ̓όβριμος ‛Έκτωρ [óbrimos Héktōr “powerful Hector”] is occasionally found, however, because it contributes the added metrical flexibility provided by an epithet that begins with a vowel rather than a consonant). Given a particular metrical space to fill, the poet was not expected to create a new epithet ex nihilo; he did not even have to pause to consider a choice between two or more inherited epithets; only one epithet was available for that metrical space.

Admittedly, noun–epithet combinations and formulaic phrases are a feature of many later epics: Vergil’s Aeneas Anchisiades; Dante’s del magnanimo quell’ ombra; Spencer’s sweet-bleeding myrrh and warlike beech; Milton’s flow’ry dale of Sibma clad with vines; and even Joyce’s snot-green sea. Their prevalence is a natural result of trying to imitate an epic style that goes all the way back to Homer. But the imitation is only skin-deep: none of these later “literary” epics has an underlying formulaic system characterized by complexity and economy. In Vergil’s Aeneid, for example, the hero Aeneas is called both pius Aeneas (˘˘/ – – / –) and pater Aeneas (˘˘/ – – / –), metrically equivalent epithets, and therefore uneconomical. Vergil’s choice between the two epithets was motivated by the context – whether Aeneas is acting piously or in a fatherly manner – rather than by metrical considerations. Indeed Aeneas is called bonus when showing kindness, magnanimus when acting bravely, and heros when demonstrating heroic qualities. Achieving le mot juste was a pressing consideration for Vergil – but not for Homer.

This system of epithets was a component of the more elaborate, and equally systematic, verse-long formulae that are so essential to Homeric diction. For example, the poet relied on the following system of formulae when expressing the action: epithet(s) + man/god/woman/goddess + answered/addressed + him/her/them:

–	– / – ˘˘/ – ˘	˘ / – – / – ˘˘/ – ×
		μέγας κορυθαίολος ‛Έκτωρ
		mégas koruthaíolos Héctōr
τòν		great shining-helmed Hector
tòn
him	δ’ ἠμείβετ’ ἐ′πειτα	Ποσειδάων ἐνοσίχθων
	d’ ēmeíbet’ épeita	Poseidáōn enosíkhthōn
τη`ν	then answered	Poseidon the earth-shaker
tē`n
her	δ’ αu̓˜τε προσέειπε	περίφρων Πηνελόπεια
	d’ aûte proséeipe	períphrōn Pēnelópeia
τοu`ς	in turn addressed	prudent Penelope
toùs
them		θεα` γλαυκω˜πις ̕Aθήνη
		theà glaukôpis Athē´nē
		goddess owl-eyed Athena
		etc.

Elaborate systems such as these suggested to Parry the traditionality of Homeric diction: no single poet could have been responsible for such systems; they had been created and nurtured by generations of bards who had passed them down from father to son, from master to student, until they reached Homer in the late eighth century BCE. Virtually every word in the Homeric epics is formulaic and traditional. Very rarely was there a need to turn a new phrase or create a new scene. The poet rather tapped into the vast reservoir of traditional diction that he had inherited.

But proof of Homer’s orality came unexpectedly when Meillet circumspectly invited Matija Murko, an expert on Yugoslavian heroic poetry from the University of Prague, to attend Parry’s defense of his doctoral thesis. Murko pointed out that the orally composed heroic poetry of Yugoslavia had the same type of traditional phraseology, operating in much the same way, as Parry was describing in Homer. Here, in the decasyllabic verse of Serbo-Croatian epic:

1 2	3 4 /	5 6 7 8 9 10
		Miloš čobanine
		Milosh the shepherd
	njemu	nahod Simeune
	to him	the foundling Simeon
veli	njojzi	Todore vezire
said	to her	Theodore the high counselor
	njima	Kraljeviću Marko
	to them	king’s son Marko
		srpski car-Stjepane
		Serbian emperor Stephen
		etc.

Here, then, was a living oral tradition that could be observed and studied first-hand, a sort of laboratory test of Parry’s theories about Homer. This was a project on which Parry was soon to embark: two long trips in the early 1930s through the mountains of Yugoslavia, where he and his assistants, Albert Lord and Nikola Vujnović, recorded over 13,000 songs on 3,500 aluminum disks and transcribed many others. These are the disks and transcriptions that make up the collection that remains one of the crown jewels of Harvard’s Widener Library to this day: The Milman Parry Collection of Oral Literature. Parry would conclude in his highly analytic way that: a) The diction of Yugoslavian heroic poetry is oral and traditional; b) The diction of Greek epic poetry shows many of the same features that in Yugoslavian poetry are due to that oral and traditional nature; c) Greek epic poetry must be oral and traditional. Thus was born the “oral-formulaic theory,” which properly speaking is not merely a theoretical construct but an empirical fact.

The practice of comparing features of different oral traditions opened up an entirely new field of scholarship. Although Parry died tragically at the age of 33, his long-lived apprentice Albert Lord extended the oral-formulaic theory to several other traditions, first to the national European epics, which had for a long time been treated as written texts: the Old French Chanson de Geste (including Roland), the Medieval German Nibelungenlied, the Hispanic Poema de Mio Cid, the Old English Beowulf; and then old Irish, Arabic, Chinese, Latvian, and Norse traditions. More recently the oral-formulaic theory has been extended to relatively modern traditions: Modern Greek, Turkish, Tunisian, Russian, African, Kurdish, Irish, Native American – to well over 100 independent language groups – and even to such genres as American folk preaching, sports broadcasting, and musical improvisation of various types, such as plainchant, ballad, folksong, blues, jazz, fiddling, and rap.

The result is that Homer’s Iliad and Odyssey have to some extent been disengaged from the genre of monumental literary epics like those of Vergil, Dante, and Milton, and positioned instead within the genre of oral epic traditions in which they more comfortably belong. One can perhaps appreciate the irony in all this: that it was by virtue of very intensive, old-fashioned, text-oriented philology that Classicists began to think of Homer as oral rather than literary.

3. Internal Evidence for Orality in Homeric Epic

This is something that Classicists should have recognized much earlier. There are many indications within the narratives of the Homeric epics themselves that their composition, performance, and transmission was oral rather than literary in nature. The first indication is that other than the mysterious σήματα λυγρά (sē´mata lugrá “baleful signs”) in the story of Bellerophon (Il. 6.168–70) – a tale drawn wholesale from the Near East, where writing had been known for millennia – the Homeric epics make no reference to writing. This apparent absence of any familiarity with the technology of writing correlates well with what we know archaeologically about the absence of any writing system during the very period in which heroic epic flourished: between the twelfth and the eighth centuries BCE.

Another indication of orality has to do with the portrayal of the bard and his audience within the epic tradition. The Iliad and Odyssey portray the epic-making process as an oral performance of a song: the picture of the two ἀοιδοί (aoidoí, “bards”) in the Odyssey, Phemius and Demodocus, is of two professionals who sing their songs to the accompaniment of a cithara or phorminx (lyre-like stringed instruments). Demodocus appears to be attached permanently to the palace of Alcinous on the island of Scheria, while Phemius seems to have taken up temporary residence in Odysseus’ palace in Ithaca. They are both honored above all mortals and thought to be divinely inspired because of their ability to sing. They have immense power in being able to immortalize the deeds of men by recalling them in their songs, or contrarily to assure that the deeds of men are forgotten by failing to make mention of them. Both bards are dependent upon the palace for sustenance: they literally “sing for their supper.”

Several scenes in the Odyssey provide glimpses of the setting and audience of bardic performance. Phemius sings, to the accompaniment of his cithara, after the suitors’ have feasted at the palace in Ithaca (Od. 1.144–55). A “divine singer” plays his phorminx during a wedding feast in Sparta, while the people dance and two acrobats revolve around them (Od. 4.1–19). Demodocus, the blind bard, who is sitting on a chair in the midst of the feast, is stirred by the Muse to sing and play his phorminx after a sacrifice by the Phaeacians (Od. 8.40–106). Demodocus sings a final song at Odysseus’ request following a feast in Alcinous’ palace (Od. 8.469–545), after which Odysseus praises the singer with generous words:

Surely it is a good thing to listen to a singer like this one, whose voice is like the gods; for I think there is no occasion more pleasant than when merriment holds sway over all the people, and those feasting throughout the house sit in order and listen to the singer, and beside them the tables are loaded with bread and meat, and the wine steward draws wine from the mixing bowl and carries it around filling the cups. This is in my opinion the very best of occasions (Od. 9.3–11).

Surely this is a portrayal of the status of the bard in Homer’s own time. In fact, Homer seems to be drawing a self-portrait here, probably for protreptic purposes. His bards are always presented in a good light: Phemius is one of very few spared Odysseus’ vengeance; Demodocus’ performance earns him the choicest cut of meat. Homer describes Demodocus as blind: “The Muse gave him both good and evil. She deprived him of his eyes, but she gave him sweet song (Od. 8.62–5).” This too may be a self-portrayal. In the “Homeric” Hymn to Apollo (169–73) a stranger is imagined asking: “Who is the best of the bards?” The answer: “A blind man who dwells on Chios; his songs are the best.” This geographical detail as well may be a self-portrayal by Homer, who in several of the most ancient traditions is said to have come from the island of Chios in Asia Minor.

A third indication of the orality of epic verse within the epics themselves has to do with the perceived status of the Muses in the process of epic verse making. Before beginning any long stretch of song, the poet invokes the divine aid of the Muses, who, as daughters of Memory, enable the poet to remember the details and the overall path of the song. As he begins the monumental task of singing the 12,110 verses of the Odyssey Homer invokes a single Muse, a goddess, a daughter of Zeus, asking her to tell the tale of Odysseus’ return, and to begin it at some point of her choosing (Od. 1.1–10) – she chooses to begin the story at the end, in the tenth year of his return, by means of a clever hýsteron próteron. Before beginning the monumental task of singing the 15,693 verses of the Iliad, Homer invokes the goddess, i.e., the Muse, to sing the wrath of Achilles (Il. 1.1–7); soon thereafter, when he is faced with the daunting task of recalling the long catalogue of Achaean ships that had sailed to Troy, Homer invokes the Muses, the daughters of Zeus, goddesses, who live on Olympus, to help him remember, for they know all things, while men only hear rumor and know nothing (Il. 2.484–93). The singer appears helpless without the Muse.

A fourth indication of the orality of epic verse is the acknowledgement within the Homeric epics, which are themselves our earliest surviving literature, that there already existed epic songs of various sorts that were broadly known. The most notorious example is Circe’s reminder to Odysseus of the ship Argo “which is on all men’s minds” (Od. 12.55–72), but there are also references to earlier tales about the Calydonian boar hunt (Il. 9.529–99), about the assault of the seven heroes against Thebes and the continuing conflicts in the next generation of these heroes (Il. 4.370–410, 5.800–808), and about several other heroes of previous generations, such as Nestor, in his youth (Il. 7.132–57, 11.669–760, 23.629–42), Heracles (Il. 2.653–70, 5.628–54, 8.362–5, 19.95–133; Od. 11.617–26), and Oedipus (Od. 11.271–80). Based simply on allusions in the Iliad and Odyssey, we can infer that Homer was familiar with at least the following epic traditions, and that they may even have been part of his own repertoire: “Jason and the Argonauts,” “The Calydonian Boar Hunt,” “The Seven against Thebes,” “The Epigoni,” “The Heraklea,” “The Nestorea,” and “The Oedipodea,” as well as various epics about the Achaean expedition against Troy and its aftermath. And these must be just the tip of the iceberg. To complement these references in the Homeric epics themselves, we have the evidence of a few fragments of later “Cyclic” poems that must be derived ultimately from pre-Homeric forms, we have some later prose summaries of these same Cyclic poems, we have several surviving tragedies, and many more names of tragedies, whose plots were based on pre-Homeric and extra-Homeric epics, and, finally, we have hundreds of depictions of scenes from these epics in the plastic arts, particularly in archaic vase painting. All of these epics, being pre-Homeric, were certainly composed, performed, and transmitted orally, since they flourished during a period when the Greeks had no writing system. In fact many of them failed to survive precisely because they were never written down. This is all a salutary reminder that although from a modern perspective Homer is commonly regarded as the father, the progenitor, of western literature, from an ancient perspective he is the son, the progeny, being the inheritor of a long epic tradition.

4. From Oral Performance to Written Text

All ancient Greek epic verses that have survived, no matter how oral their background, eventually became written texts; had this not occurred, we would never have known them. When, where, why, and how they became texts are very contentious matters, however, and much ink has been spilt over these questions during the past few generations.

One view is that the epic poets themselves learned how to write and took advantage of this new technology to record their verses in a more fixed and stable medium. Advocates of this view thus account for some of the apparent features of literacy in post-Homeric epic verse, such as Hesiod’s Theogony and Works and Days, as well as the “Homeric” Hymns, but they also attribute the extraordinary length, structure, and monumentality of the Iliad and Odyssey precisely to Homer’s ability to write. Some have even suggested that the alphabet was adopted by the Greeks specifically for the purpose of recording the Homeric epics. This view is perhaps motivated by some wishful thinking: it avoids the messiness and complication of other explanations; it offers a romantic notion of an individual poet and his text with which we as literates have become familiar and comfortable; it allows an editor to strive to reconstruct the original text of Homer in much the same way as he would the original text of Apollonius of Rhodes. But it fails to account for many features in the Homeric texts that indicate that they were not slowly and deliberately written down, with the leisure to reread, reconsider, and revise (see further below). Moreover, it seems most unlikely that it would ever have occurred to a truly orally composing singer to write his songs down as a text. Preservation of the exact words of his songs was never his goal; he could perform these songs again at any time. A written text served him no purpose in performance; on the contrary, he probably performed more freely and comfortably when allowed to sing at his own pace to the accompaniment of his lyre, not with an unwieldy text to encumber him. It must have been someone other than the singer who came up with the idea of recording the songs as written texts.

A second view, the polar opposite of the first, is that the textualization of the Homeric epics was a long and complicated evolutionary process: that the epics remained largely oral, and therefore fluid and unstable, not only among the bards of the earliest period, but also among the rhapsodes of the Archaic period, and even into the Classical and Hellenistic periods in disparate local traditions, well after at least some degree of stabalization was achieved by the textualization of a Panhellenic version at the Panathenaic festival in the late sixth century BCE. The Homeric epics did not become the texts as we (more or less) know them until the Alexandrian librarians of the third and second centuries BCE standardized and canonized them. This view tends to efface Homer’s existence as a human being and instead attributes the Iliad and Odyssey to a mythic figure, a cultural icon, a symbol of oral tradition that we can call, for the sake of shorthand, “Homer.” But the epics were actually shaped by generations of mouths and hands, slowly crystallized, and not really fixed until the late Classical or even Hellenistic period. This evolutionary view is attractive in many respects, since it offers an explanation for several curious developments relevant to the transmission of the epics: the relative paucity of depictions of Iliadic and Odyssean scenes in the graphic arts during the Archaic period, followed by a surge in popularity of such scenes in the late sixth century BCE (i.e., coincident with a Panathenaic textualization); the sometimes remarkable differences between our inherited texts and the quotations of Homer by Classical authors, the textual versions reported to have existed in the manuscripts available to the Alexandrian editors, and the longer and “eccentric” readings of the Ptolemaic papyri; and the late linguistic forms, especially the “Atticisms” and “hyper-Ionisms” that reside, at least on the veneer, of our inherited texts. But the drawbacks of this view are numerous as well. It fails to account for many important features of the Homeric texts: the overall unity of their narratives; the various types of inconsistencies that remain embedded in their narratives; the absence of multiple versions of the Iliad and Odyssey; and the fact that the development of the epic art-language appears to have been arrested at a particular moment in time (see further below). As a practical matter the evolutionary view imposes nearly impossible challenges on the modern editor of the Homeric texts, for all textual variants are regarded as potentially authentic readings. How is the modern editor to present the fluidity and multiformity of the epic tradition in the form of an edited text that has conventionally placed readings of a supposed original in the favored position above, while demoting supposed variants to the level of the apparatus criticus below?

A third view, which falls somewhere in between the other two, though much closer to the first, is that Homer dictated his songs to a learned scribe (or scribes), who recorded the words, probably with a reed pen on papyrus (or perhaps leather). The idea of textualizing the songs did not come from the singer, who, as we have noted, placed no value on written text. It must instead have come from a patron, a sponsor, or a simple admirer, who was familiar with the only mechanism capable of accomplishing this task: oral-dictation. Since the alphabet with which the songs were first textualized originated in Phoenicia, and since the papyrus upon which the text was first transcribed originated in Egypt, and since many components of the songs themselves – the tale-types, themes, and poetic forms – originated in the Levant and Mesopotamia, it does not require too great a leap of faith to suppose that the very idea of writing down the songs originated from someone acquainted with the civilizations of the Near East, where the writing down of epic songs, some even through the process of dictation, and their transmission by means of written text, had been practiced for more than a millennium. Our inherited texts of the Iliad and Odyssey, in this view, are more or less reliable records – though passed through countless hands over many generations – of what was once an oral-dictated text, i.e., a scribal transcription of a performance orally delivered by a historical Homer in the eighth century.

One obstacle to this view is that the scenario may seem a bit hard to imagine: an oral poet, accustomed to performing his songs to the accompaniment of a lyre to a live audience, is persuaded to perform, or dictate, his songs, probably at a much slower pace than usual, to a scribe, or group of scribes, who record the words of his performance as written text. Yet there are analogues to this process in comparative oral traditions: Parry imagined Homer dictating his songs, while someone else with writing materials wrote them down verse by verse, much as the Serbo-Croatian guslari whom he was observing dictated their songs to his assistant Nikola Vujnović (Parry 1971, 451). Albert Lord made further comparisons, and gave the oral-dictation theory a clear and thorough articulation (Lord 1953; 1960, 38–9). The benefits of this view are many: Homer remains a truly oral poet, but at the same time a particular text can be ascribed to him; our inherited texts of the Iliad and Odyssey can be regarded as vestiges of a real historical performance, delivered at a particular time, in a particular place, by a real living person; several of the most serious obstacles to the other views are obviated, namely that the texts that we have inherited retain several features that seem utterly incompatible with the view of an evolutionary process and, at least in some important respects, with the view of a literate Homer. I wish to highlight four of the most prominent of these features: the unity of the narratives; the various levels of inconcinnities that remain embedded in the narratives; the absence of multiple versions of the Iliad and Odyssey; and the fixation in time of the epic art-language.

The Unity of the Narratives

The Iliad and Odyssey are unified narratives whose structures are most clearly observed, whose plots are most intelligently followed, and whose nuances are most pleasurably appreciated, whether by ancient listeners or modern readers, when experienced in their entirety and within a limited time-frame. Episodes are organized in a thoughtful sequence from beginning to end and bound together by a network of interconnected references, by anticipatory and retrospective allusions, by comparative and contrasting parallelisms, and many other similar structuring devices. Neither epic is simply a collection of loosely related episodes – which would be the predictable result of a process of compilation by various hands over a long period, or of a process of gradual accretion within an impersonal oral tradition. Each epic is a work carefully arranged by a personal and inspired singer composing in a performance that was experienced in toto on some occasion that provided considerable leisure: a festival, perhaps, or a nobleman’s funeral or wedding.

Various Levels of Inconcinnities Embedded in the Narratives

The Iliad and Odyssey have survived to our day as texts that, even in the forms that have been copied and recopied for many generations, do not have the appearance of having gone through an extensive editorial process – proofreading, correcting, reworking, etc. On the contrary, they retain many features typical of oral composition-in-performance that once uttered could not be retracted. Indeed inconcinnities remain embedded in every level of our inherited texts as vestiges of their origin in oral performance: metrical blunders attributable to the pressures of oral composition-in-performance by a singer who did not go back to his verses after his performance to tidy up the prosodic loose ends (e.g., prosody of Od. 7.89); dictional inconcinnities that have resulted from stock formulaic phrases being used in contextually inappropriate circumstances, and whose survival in our texts show that the poet had no opportunity or desire to summon back his words or revise them (e.g., the corpse that “groans heavily” at Il. 13.423); small factual errors and larger narrative anomalies that point to a one-time oral-dictation of an epic composition-in-performance that was transmitted thereafter, blemishes and all, with remarkable faithfulness in its textual avatars (e.g., the Trojan soldier Melanippus, who is killed three times over a nine-book stretch of narrative [Il. 8.273–7, 15.572–84, 16.692–7]). These are not normal features of deliberately written texts, nor are they conceivable in the evolutionary model; they arise rather from the exigencies of live oral performance which, on the one hand, require that the singer extemporize as he composes during the very act of performance, and, on the other hand, prohibit the singer from retracting or correcting his song once it has left his mouth.

The Absence of Multiple Versions of the Iliad and Odyssey

The evolutionary model, hypothetical in the case of the Homeric epics, has been applied appropriately and productively to the presumed – and in some cases demonstrable – histories of several other oral epic traditions that were eventually fixed in textual forms: e.g., the Sumerian and Akkadian versions of the Epic of Gilgamesh, the Sanskrit Mahabharata and Ramayana, the Persian Shahnama. It is instructive, then, to compare the features of the surviving texts of these other epic traditions with those of the Homeric tradition in order to determine if the (hypothetically) similar circumstances during the composition and transmission of these documents have produced similar outcomes. What we discover is that in these other epic traditions there have survived multiple versions that are substantially different from one another, not only in small matters of diction and detail, but also in their essential poetic forms, their larger themes and narrative patterns, their overarching plot structures, and even their total lengths. These multiple versions all have equal claim to authenticity; hence, the search for an archetype is meaningless. In the case of the Iliad and Odyssey, however, multiple versions have not developed.

While it is true that textual variants occur in quotations of “Homer” by later Classical authors of the fifth and (mostly) fourth centuries BCE, in the reports of the Alexandrian scholars of the third and second century BCE about what they read in earlier editions of Homer, in the readings of the 40 or so surviving remnants of Homeric texts on papyri from the early Ptolemaic period (third–second century BCE), and, though to a much lesser degree, in the 900 or so surviving manuscripts of the post-Aristarchean “vulgate” (i.e., after around 150 BCE), from the perspective of the monumental epics as a whole these variants are comparatively trivial and do not provide the evidence for substantially different versions of the Homeric texts. We have only one version of the Iliad and one of the Odyssey, with the same characters, the same story, and even the same sequence of episodes – all of which are, moreover, told in a very uniform meter, dialect, diction, and style throughout. There is no evidence that there ever existed any texts of Homer’s Iliad without a Patroclus, or of Homer’s Odyssey without a Telemachus. Nor is there any evidence of texts of the Iliad and Odyssey that were half the size, or twice the size, of our inherited texts. It seems likely, then, that, unlike these other epic traditions, our Iliad and Odyssey each go back to a single archetype that was fixed in writing and whose text did not thereafter suffer substantial editorial tampering.

The Fixation in Time of the Epic Art-Language

Our inherited texts of the Iliad and Odyssey reveal a language that was frozen in time, a language that had previously been evolving hand in hand with the vernacular but that had in its eighth-century Ionic form become fixed. There had once existed a vibrant Mycenaean epic tradition, but our inherited texts are not Mycenaean (though there are Mycenaean words and phrases, even poetic formulaic phrases, embedded in them); thereafter there had existed a vibrant Aeolic epic tradition, but our inherited texts are not Aeolic (though Aeolic words and phrases abound, especially ones that provide metrically useful alternatives to the corresponding Ionic forms); thereafter there arose a vibrant Ionic epic tradition, and this is when the linguistic evolution that had so characterized the epic tradition previously was arrested. Though the epics continued to be performed and enjoyed – recited orally and received aurally – the epic Kunstsprache “art-language” in which they had for so many generations been composed had become a “dead” language. The language of both the Iliad and Odyssey attained a high degree of fixation precisely at this period, substantially in the Ionic dialect, and they continued in their later transmission to retain their Ionic forms. This fixation was surely due to textualization. Whether the writing down of these epics enabled them to gain an exceptional status, or whether an exceptional status caused them to be written down, it was textualization, the result of oral-dictation and transcription at a specific time and place (in the case of the Iliad and Odyssey during the eighth century BCE in Ionia), that assured linguistic fixation. For the epic language did not continue to evolve linguistically – to create innovative forms and formulae – through the seventh, sixth, and fifth centuries, and thereafter, as it had in its more fluid oral form before the eighth century. The so-called “Atticisms” and “hyper-Ionisms” that occur occasionally in our inherited texts are simply a veneer: metrically equivalent modernizations and modifications of an already established text.

5. Homeric Epic as Written Text

We have been following the transformation of the Homeric epics, beginning with their truly oral stage, when they were orally composed, performed, and transmitted by an ἀοιδός (aoidós “bard”) singing to the accompaniment of his lyre. Thereafter the epics went through a stage during which they were no longer orally composed or transmitted, though they continued to be orally performed by ῥαψῳδοί (rhapsōidoí, “song-stitchers”) before a live audience, through recitations of memorized texts, albeit with a certain amount of manipulation and extrapolation. The so-called Homeridae, “descendants of Homer,” a guild of rhapsodes from Chios from the sixth or possibly seventh century BCE, fall into this category (see Pindar’s Nemean 2.1–3 and scholia thereon; also Plato’s Ion 530d, Republic 599e, Phaedrus 252b). The third stage was completely textual, when the Homeric epics began to be thought of as collections of papyrus rolls to be bought from a bookseller and perhaps even stored in a library. Athens enjoyed a commercial business in the selling and buying of books by the late fifth century (Eupolis F 327 [PCG]; Plato’s Apology 26 d–e), and Plato’s Academy and Aristotle’s Lyceum had actual libraries for research purposes. Yet, even during this stage, the Homeric epics continued to be read aloud, usually in a public forum, as was the common practice with other literary texts as well.

The fascinating history of the Homeric epics did not cease when they became texts (cf. Armstrong, ch. 2, pp. 36–8 in this volume). I wish to conclude with a few words about how these once oral epics became the most literary of literary texts during a particularly bookish period of Greek history: the late Hellenistic period (cf. Nünlist, this volume, ch. 19, pp. 301–7). This was the period of the great scholarly libraries of antiquity: Alexandria, Pergamon, Antioch, etc (cf. Hose, this volume, ch. 21, pp. 331–5). The Homeric texts from around the world were collected, collated, and edited. Each epic was divided into 24 books, and the verses of each book were numbered. The text was weighted down with several types of punctuation marks, along with occasional aspirations and accentuations. Verses were annotated with various critical signs that indicated their perceived textual status. Lexica, glossaries, and commentaries were produced. Scholars debated fine points of textual criticism, word division, and etymology.

The lexical activities of scholars were matched by those of students. Homer’s epics had occupied a privileged position in the educational system since the early Classical period (see Plato’s Republic 376e–98b; Plutarch’s Alcibiades 7.1). Once students had learned the basics of reading and writing – the alphabet and the syllables – they proceeded at once to Homer, which they read, recited, and memorized. We hear of a certain Nicaratus, who claims that his father made him memorize the entire Iliad and Odyssey (Xenophon’s Symposium 3.5). This attention to Homer continued to be the norm during the Hellenistic period, during which time education throughout the eastern Mediterranean world became remarkably homogeneous, composed of a standard ἐγκύκλιος παιδεία (enkúklios paideía, “cycle of education”) that had Homer as its centerpiece. Lists of Homeric vocabulary were used to learn letters, syllables, and then full words, beginning with monosyllabic words and working up to pentasyllabic words. Homeric verses, especially those of a gnomic nature, were used as models for writing exercises. In the 400 or so school texts that have been discovered among the Egyptian papyri of the Hellenistic and Roman periods are found nearly 100 quotations or citations of Homer. Homer is quoted five times more often than the second most popular author, Euripides, and thirteen times more often than the third most popular author, Menander. Those who continued their education beyond the primary level were expected to know the entirety of the Homeric epics, not simply for their poetic value, but also as a source of mythical, moral, philosophical, geographical, rhetorical, and grammatical information. We cannot help but be struck by a certain irony here: that these truly oral songs that were once composed, performed, and transmitted without the use of writing ended up as Hellenistic school texts that were studied by children to learn their ABC’s.

REFERENCES