Homer and the Definition of Epic

Epic, as a genre, is defined using many different criteria, from mode of discourse (although some epics are not predominantly narrative), length (though some epics are short), relationship to other genres (though not all epics incorporate minor genres), subject matter (though not all epics involve war or travel), theological framework (though not all epics mention the gods), national or ethnic significance (though not all epics are closely linked to a particular nation or ethnic group), elevation of diction (though not all use high language), mode of composition (though not all epics originate from oral composition), mode of dissemination (though not all are primarily intended for oral performance), and metre (though not all use the dactylic hexameter). [1] It seems that one of the very few issues on which all agree – perhaps the only one – is that the Homeric poems are examples of the genre. This paper aims to explore the place of Homer in the vast and multiform world of epic poetry. For obvious reasons, the discussion has to be selective: I begin by looking at some fundamental tensions in current approaches to Homeric poetry within the field of classical scholarship; I then trace the origins of some of those tensions to the work of Milman Parry and show how his comparative research had major consequences not only for the study of Homer but also for the definition of epic, and of literature more generally. I then illustrate my point by focusing on a complex and heated debate over the existence and nature of epic poetry in sub-Saharan Africa. [2] Finally, I return to Homeric poetry and ask how even a brief foray in the wider world of epic affects our interpretation of the Homeric poems and our sense of their place in the present literary and political landscape.

Anyone even superficially acquainted with the field of Homeric scholarship knows that it is characterised by lively, even acrimonious debate, rather than by insipid agreement. When, as a Masters student in Oxford, I was trying to formulate some ideas for a Ph.D., I bumped into a famous Homeric scholar in the street and, quite inappropriately, sounded out some of these ideas with him. His only comment was: ‘Well, and then there is Nagy, and what he’s doing with Homer is dangerous, very dangerous…’ It was a shameful moment for me, because I had no idea what he was talking about. Worse still, I did not dare ask him: I nodded solemnly and crossed the street. Ten years on, I am still not entirely sure I can account for quite the level of fear and alarm I witnessed on that occasion, but perhaps the mysterious incident can be put in a wider cultural context. My aim, here, is to show that the interpretation of Homer has important implications in the first place for the definition of epic, but ultimately for the very conception of ‘literature’.

As Ford points out,

I return to Aristotle’s Poetics in section 3, but for the time being it may be useful to investigate further the place of the Iliad and the Odyssey in the cultural landscape outlined by Ford: it seems that Homer is presented here simultaneously as the most central author of the western literary canon and as sharing fundamental similarities with a vast range of epics which are deemed neither literary nor western. Now, Ford’s phrasing reveals an obvious unease with this dual definition of epic (‘with equal justice’, ‘for all their divergence’) and goes on to point out at least one point of contact between these two ‘quite different sets’ of epics, namely Aristotle’s Poetics. There are, in fact, many more points of contact: in a collection of articles that Emily Greenwood and I are hoping to publish, we explore precisely the connections, mutual influences and common ground between different approaches to Homer and hence epic more generally. For example: in some histories of literature, Gilgamesh rather than the Iliad is presented as the ultimate source and origin for the western epic tradition; or again: the novelist Ismail Kadare uses the oralist’s Homer to claim a place for himself and Albania in the literary salons of Paris; or, perhaps most famously, Derek Walcott presents himself simultaneously as the heir of Milton and of a vibrant oral tradition.

For the time being, however, it may be useful to maintain Ford’s dichotomy as a heuristic starting point, not least because it seems to me accurately to reflect some problems in current Homeric scholarship. Some classicists study Homeric epic together with other examples of what is standardly called oral/traditional poetry. In a recent book, for example, John Foley pairs Homer with a Tibetan Paper-Singer, a North American Slam Poet and a South-African Praise-Poet, in an attempt to work out ‘how to read an oral poem’. [4] Other scholars, by contrast, insist that Homeric epic should not be viewed as a ‘primitive, unidimensional artefact’ but rather be seen as the first great work in a long and distinguished genealogy of western literature. [5] That tradition, needless to say, does not include Tibetan Paper-Singers. It thus seems that Homer tends to be paired either with Virgil or, say, with Avdo Međedović, the super-star of muslim Serbocroatian epic. Depending on the company he keeps – and this is my main point – he considerably changes character: the aims and tools of Homeric scholarship depend in large measure on whether Homer is paired with Virgil or the South Slavic bards. What is more, it seems that up to now Homer’s bedfellows, Virgil and Međedović, have had little occasion to meet.

In the last ten years, we have witnessed the publication of two separate and, in many ways, self-consciously alternative Companions to Homer. In 1997 Ian Morris and Barry Powell edited for Brill the New Companion to Homer, which focused on an analysis of Homer as oral/traditional poetry, on metre and dialect, and on Homeric society. Last year, Cambridge University Press published the Cambridge Companion to Homer, edited by Robert Fowler, a volume which explicitly presents itself as a ‘literary’ study of Homeric epic. What is meant by this emerges from the chapters which focus on such topics as plot, characters and characterisation, speeches and similes. The Cambridge Companion also has a very substantial section on reception, which includes articles on the Hellenistic and Roman Homer, on Homer and Milton, on Homer and Joyce’s Ulysses. A self-consciously literary genealogy is being asserted here as it is explored, and it is precisely that genealogy that has been picked out for criticism. Barry Powell reviewed the rival Companion for Bryn Mawr and concluded as follows:

Now, the problem of how to write the reception of Homer is real enough, but the solution – to drop it in favour of ‘earlier topics’ – is not a viable response in all contexts. In order to understand the place of Homer, and classics more generally, in the present cultural and political landscape it is essential to grasp the nettle of Homeric reception. As has often been pointed out, the discipline of Classics itself rests on a particular understanding of the reception of Homer in Rome and in later European literature. Glen Most has recently summarised some basic assumptions which have shaped (and in many contexts continue to shape) Classics as an academic subject:

Leaving aside the polemic about prose and verse composition, Most is echoing here some well-established views about the relationship between Greece, Rome, and the modern European and American world. In the formulation of those views, the figure of Livius Andronicus, the third-century BC Roman translator of the Odyssey, often plays a major role. In his Geschichte der römischen Literatur of 1913, for example, Leo made the following claims: ‘Western civilisation depends on Graeco-Roman culture’; [8] ‘Roman culture is the spiritual link between the ancient and the modern worlds’; [9] ‘Roman literature is the first literature dependent on the Greeks, the first secondary, non-local literature’; [10] ‘Livius Andronicus is the beginner of the first derivative literature of our cultural universe’. [11] The history Leo outlines is one of conscious appropriation and adaptation, and Livius Andronicus is, to use a phrase proposed by Scevola Mariotti, the first exponent of ‘artistic translation’. [12] Given this overall framework, it is perhaps unsurprising that the concepts of allusion, adaptation, translation, and transference have played a particularly important role in the study of Roman epic.

Recently, however, the basic aims (and related methodology) of Latin scholarship have been fundamentally questioned. For example, in a recent collection on pre-modern epic, Von Göttern und Menschen erzählen. Formkonstanzen und Funktionswandel vormoderner Epik, Georg Rüpke argues that early Roman epic should be understood in the first place in its social and religious context, and not simply as a link in a long history of literary translations and adaptations. [13] The starting point, according to Rüpke, must be the performance of epic. In reviewing the article, Alessandro Barchiesi points out that Rüpke’s suggestions represent, in effect, a paradigm shift not only in the study of Roman epic, but also in Latin Studies more generally. Barchiesi also suggests that this shift follows what has already happened in the study of Homer:

In this recent debate about the study of early Roman epic, we thus witness an interesting paradox: precisely at the moment when Roman epic tries to emancipate itself from the influence of Homer as a literary model, Homeric poetry comes again to the fore, in this case as a means of exploring oral performance. This tension about the place of Homer in the study of epic can be stated much more generally. Western literature has often been conceptualised as a huge family tree. In his Anxiety of Influence, for example, Harold Bloom views connections between texts as Oedipal relationships between fathers and sons. [15] In his giant dysfunctional family, Homer has a place of honour. Bloom writes: ‘Everyone who now reads and writes in the West, of whatever racial background, sex or ideological camp, is still a son or daughter of Homer’. [16] This is a grandiose statement about the place of Homer in the modern world but, as we have already seen, it seems reductive in at least two respects. The first and most obvious is the restriction of Homer’s influence to ‘the West’, however conceived. The second is the genealogical conception of literary history it adopts: Homer is primarily viewed as an ancestor with many sons and daughters, [17] yet we have already seen that the world of epic poetry is often construed by analogy and comparison, rather than on a model of blood descent.

The origins of the different conceptions of epic outlined in section one can in part be traced to the ground-breaking work of Milman Parry in the 1930s. As has often been noted, the idea that Homeric epic might have originated in oral composition had been around for a long time, as indeed had the assumption that Homer shared important features with many non-literate cultures. [18] Parry’s work, however, was revolutionary, also because he packed his suitcases and tape-recorders and went to study the epics that were actually being performed in an area which is now divided between Serbia, Croatia, Kossovo, and Albania. He did so with commitment, rigour, and astonishing insight, thus making a substantial contribution to the study of Muslim Serbocroatian and Albanian epic. But his work is remarkable also for its lasting impact on Homeric scholarship. No classicist today can simply ignore Parry’s comparative approach to the study of the Iliad and the Odyssey. [19]

Parry argued that the Homeric poems were composed relying on similar techniques to those used by South Slavic bards. Both epic traditions had developed, for example, a complex and highly efficient system of formulae which enabled the bard to re-compose in performance. In studying the principle of formulaic economy, Parry explicitly drew a distinction between Homer and the South Slavic bards on the one hand, and Virgil and Apollonius on the other. For example, in his study of noun-epithet clusters, he started with the following hypothesis: ‘It might be believed that the systems of noun-epithet formulae which we have shown to exist in Homer are a common property of all hexametric poetry, being due not to the influence of metre over successive generations, but to the influence of metre on the style of poets of any period whatever. A study of the heroic poems of Apollonius and Virgil will provide us with sure evidence on this point’. [20] He then concluded his study as follows:

The technical nature of Parry’s discussion should not distract us from the fact that he was effectively driving a wedge between Homer and Virgil: the Roman poet simply did not understand, ‘or even suspect’, what considerations governed the choice of epithets in Homer. Those considerations, by contrast, became clear from a comparison between the Iliad, the Odyssey, and the South Slavic epics. The split between the world of literary history, where one poet consciously imitates the style of another, and that of oral/traditional epic, where composition techniques are in part dictated by performance context, can thus clearly be traced back to Parry’s study of Homeric versification. What is more, the kind of dichotomy between Virgil and Serbocroatian epic predicated by Parry has left a profound mark on classical scholarship. It has often been pointed out that the analogy between Homeric and South Slavic epic has encouraged but also in many ways hindered comparative approaches to epic: we now know, for example, that few oral epic traditions are formulaic quite to the same extent or in the same way as Serbocroatian and Homeric epic. [22] But Parry’s other legacy, the separation between Homer and Virgil, was equally important for the future study of the genre.

Parry saw himself mainly as contributing to a current debate over the Homeric Question, and accordingly centred his research on the issue of composition; yet his work also had momentous implications for the criticism and evaluation of Homeric epic. The question how to interpret an oral/traditional epic became pressing and fuelled an often acrimonious debate – a debate which is still very much alive. [23] From early on, it became clear that the interpretation of Homeric epic had to be radically re-thought in the light of Parry’s work. Classicists worried, for example, about the interpretation of epithets: if they were actually chosen for the help they provided in ‘the handling of names’, then it was perhaps problematic to talk about their meaning and effect in a particular line of poetry. Parry himself famously suggested that they should not even be translated. [24] The whole problem of meaning fuelled an intense discussion about fundamental issues of interpretation such as the influence of tradition, the role of authorial choice and intention, the status of the written text, the input of context and audience in the production of meaning, the criteria for aesthetic evaluation. Parry himself found it difficult to tackle some of those issues, particularly those relating to literary merit. For example, he was sensitive to the beauty of traditional formulations and was eloquent about it, while at the same time described Apollonius as the first truly creative poet. [25]

As must by now be obvious, Parry and his collaborator Lord had a major impact on later conceptions of epic and on the split between ‘Western literary epic’ and ‘oral epic to be found the world over’ (to use Ford’s labels). It seems that whereas Homer can, uneasily, belong to both epic worlds, Virgil effectively acts as a discriminator. In the Cambridge Companion to Homer, for example, two articles explicitly place Homer within a wider epic context. John Miles Foley writes about ‘Epic as genre’: his article does not even mention Virgil. Penelope Wilson, by contrast, writes about ‘Homer and English epic’ and rightly points out that Virgil’s influence in this context is determining: Homer is inescapably read through the Roman poet. It may seem strange that no connections are made, within the volume, between these two chapters on Homer and epic, but this failure is more explicable if we take into account quite how radical Parry’s vision was. We may notice, for example, that he presented Avdo Međedović and other bards as modern embodiments of Homer not long after the publication of Leo’s Römische Geschichte: his conception of Homer’s relationship to the modern world could not have been more different. The point here is that Leo’s vision of epic poetry and western culture was at least as persuasive and influential: we have seen how it reverberates through recent pronouncements on ‘moral perfection and stylistic refinement’. [26]

Unsurprisingly, given the dominant conception of the classical tradition, the legacy of Milman Parry is often assessed through arguments which insist on Homer’s literary merit. Robert Fowler, for example, sees formularity and ‘art’ as diametrically opposed:

From this premise, he goes on to suggest that Homeric epic may yet be oral, if enough examples of ‘subtlety and sophistication’ in oral epic can be found:

This statement exemplifies some of the pitfalls of Homeric scholarship: Homeric epic may be similar to traditional oral poems the world over, but only if the latter prove to be ‘equal’ to it.

If we ask what criteria are used to establish such equality, we quickly return to notions of great literature which ultimately stem from the perceived equality between Homer and Virgil. Virgil translates Homer for Rome and is in turn translated by later poets into the canonical literatures of Europe. [29] Virgil’s Aeneid implies a value judgement about the Iliad and the Odyssey and in turn demands to be measured against the model. Such acts of evaluations lie at the heart of the western literary canon, literally conceived as a ‘measuring rod’. Herrnstein Smith argues:

We are confronted, again, with our familiar dichotomy between a transcultural and a western Homer, though this time couched in the most general terms, as part of a discussion of literary merit. The interpretation of Homer, it seems, has fundamental implications not only for the definition of epic but, more generally, for the definition and evaluation of great literature.

So far, I have argued that under any definition of epic the Homeric poems are seen as examples of the genre. I have also suggested that current tensions and disagreements over the interpretation of Homer are related to larger debates about the nature of epic and literature more generally. In this last section, I aim to discuss the definition of epic by starting not from the Iliad and the Odyssey, but from a range of performances and texts which have sometimes been denied membership to the epic genre. As will soon become obvious, the debate over the existence and nature of epic in sub-Saharan Africa sheds light on many issues which also plague Homeric scholarship: for example, the relationship between tradition and individual talent, the status of the written text, the input of context and audience in the production of meaning, and the criteria for aesthetic evaluation.

The controversy has its origin in a by now infamous statement by Ruth Finnegan from her seminal Oral Literature in Africa: ‘Epic hardly seems to occur in sub-Saharan Africa’. [31] This statement immediately provoked a series of studies which set out to prove her wrong. [32] Among her critics were scholars from many different countries and disciplines, so it is not surprising to find that they adopted a number of different strategies to refute her claim. Before looking at some of them in greater detail, it may however be helpful to establish how Finnegan reached her conclusion in the first place. Although she does not state so explicitly, she clearly adopts an Aristotelian definition of epic as a genre. An epic, she assumes, must be a long narrative in verse, dealing with ‘heroic themes’, e.g., as she puts it, ‘the birth and tribulations of the hero, his travels and leadership of his people, finally his death’. [33] Holding on to this definition, she looks for examples in sub-Saharan Africa and, as it turns out, fails to find them. She thus concludes that epic is a far more circumscribed phenomenon than is sometimes assumed: far from being the natural expression of illiterate people, it is a specific type of poetry which occurs in some places but not others. She writes: ‘The a priori assumption that epic is the natural form for many non-literate peoples turns out here [i.e. in Africa] to have little support. [34] Epic, she concludes, is just one form of oral poetry among many. It is possible to read her work as an attempt to establish the individuality of African oral literature and the limited application of Aristotle’s definition of epic beyond its local context, but this is not how she was read. Rather, she was seen as excluding several African traditions from a category of analysis, provided by the term ‘epic’, which seemed relevant and useful to many scholars, practitioners, and audiences of those traditions. [35]

Refutation of Finnegan took many forms. Some critics accepted her framework but questioned the extent of her knowledge of African literature: the Belgian anthropologist Biebuyck, for example, argued that the Nyanga epics performed in what had then recently become Zaire, ‘admirably fit the standard definitions and characterizations of epic literature’ provided by Bowra, Chadwick and Chadwick, and Finnegan herself. [36] Particularly important in trying to refute Finnegan’s notion at a factual level was the analysis of the story of Sunjata, performed extensively in Western Africa in various Mande languages. Today, Sunjata is acknowledged as ‘one of the world’s greatest living epic oral traditions’ and has been published, for example, in the Penguin Classics Series. [37] In 1970, however, when Finnegan’s book appeared, the only available versions of Sunjata were published in prose and therefore instantly excluded from the category of epic which, according to Aristotle and Finnegan, had to be in verse. However, the problem did not just stem from the contingencies of available printed versions but from the very notion of verse. Partly as a response to Finnegan, the American linguist Charles Bird demonstrated that in the Mandekan languages (in which the Sunjata epic is widely performed), accent is generally dependent on tone, but not in ‘poetry’ where ‘accented syllables conform closely to the musical beats, and, in many cases, linguistic accent will be shifted to conform to the musical requirements’. [38] In short, the rhythm of the words in the performance of Sunjata would mark them off as different from ordinary speech, but the difference would not necessarily be captured by transcriptions using the alphabet alone. These basic issues fundamentally affected debates over the existence and nature of African epic. But the problems were not limited to empirical debates about whether a particular performance should be understood as poetry, song, or prose: the issue of orality affected, and continues to affect, the overall conceptual framework of the debate.

In a recent study of the enanga tradition of the Bahaya people of Tanzania, Mugyambuso Mulokozi discusses Finnegan’s claims from a conceptual point of view and summarises what he sees as Finnegan’s shortcomings as follows:

Mulokozi’s objection seems at first sight devastating: in a book on oral poetry, Finnegan draws her definition of epic, at least implicitly, from Aristotle, a Greek thinker who was actually peculiar in his lack of interest for oral performance. [40] And yet, one may argue, the term ‘epic’ is after all Greek in origin, and does seem to belong to an Aristotelian tradition of criticism which focuses on texts rather than performance contexts. Classicists may thus be tempted to turn the tables and direct Mulokozi’s criticism against his own work. By insisting that the Haya enanga tradition, among others, should be described as epic, is he not himself ‘imposing alien criteria deriving from a written tradition’? The issue seems to turn not just on the meaning but on the power and, ultimately, the usefulness of the term ‘epic’. We may well ask why so many scholars insist that the term epic is an appropriate description for the African traditions they study. At a very broad level, one can simply say that the performances labelled as epic thus enter a different context of fruition: they start to be discussed in Universities, international conferences, and books. But one can be more specific than that: after all, they may (and in many cases do) enter into academic discourse under other descriptions such as, for example, ‘song’, ‘myth’, ‘folklore’, ‘performance art’. We must therefore ask ourselves what makes the term ‘epic’ a useful category for the many scholars who choose to use it.

There are many answers to this question and even a superficial understanding of the whole debate immediately reveals a plethora of different agendas. Yet one important strand in the debate is the wish that African oral literature may be viewed as precisely that, literature. This is how Isidore Okpewho opens his book on The Epic in Africa, published by Columbia in 1979:

Okpewho goes on to point out that the study of African art has been subordinated to that of ritual and religion. This approach, he continues, has made it difficult to talk about the ‘aesthetic principles on which the art was executed’, about the ‘songs and stories whose primary intent is to entertain’, about the proverbs that are ‘outstanding more for style than for content’. [42] Far from looking for Levi Strauss’s ‘deep structure’ or Durkheim’s ‘collective conscience’, he argues that we should focus on the surface, on the choices operated by the bards and the criteria used by audiences to judge them. Okpewho is eloquent about anthropological categories – dirty jokes which tend to turn into fertility rites – and observes that the presence of a foreign observer who, pen in hand, intently take down performances may inhibit critical comments on the part of expert audiences: why spoil the fun of the foreign guest. [43] But his objections to anthropological approaches are, in fact, more fundamental:

It is above all in the study of the expert, creative manipulation that the category of epic, together with the methodology and critical discourse associated with it, proves to be useful. In other words, a discourse about literary merit and discrimination between excellent and mediocre epics is exactly what is wanted.

Okpewho, to be sure, has been criticised by some scholars for his excessive interest in, and reliance on, classical literature and scholarship. Johnson, for example, writes: ‘Okpewho […] compares practically every critical concept to the Greeks. He frequently mentions ideas related to context and performance […], but he fails to demonstrate how these are translated into action by the bard’. [45] Yet Johnson himself finds it important to use the category of ‘epic’ for Sunjata and many other traditions, and his reasons are not entirely dissimilar from those put forward by Okpewho: he too is interested in discussing ‘plot’, ‘narrative structure’, ‘aesthetic tension’, and the ‘literary purposes’ which lead individual bards to produce their own versions. [46] Likewise Mulokozi, in a recent and ambitious discussion of epic in Tanzania, expresses an interest in the contribution of individual bards and the merits of their work as judged within the context of performance.

It seems, then, that we have come full circle. Latinists like Rüpke and Barchiesi argue that we should pay more attention to the social and religious context for the performance of Roman epic; meanwhile Okpewho, Mulokozi and others maintain that African epic demands sophisticated literary criticism able to tackle the most fundamental issues of aesthetic evaluation both within a local context and in a more general debate about epic. We may well ask where all this leaves us as far as Homer is concerned.

One important aspiration on the part of Biebuyck, Bird, Okpewho, Johnson, Mulokozi, and several others is that of creating a broad category of epic which embraces several local traditions, among which Greek epic is but one. [47] Johnson discusses the effect of this shift on the place of Homer in contemporary culture:

In this vast and multiform world of epic poetry, Homer seems smaller and, in important respects, deficient: we cannot hear the sound of his lyre. Apart from defamiliarising early Greek epic, the wider world of epic poetry helps us to focus on the fact that the very term ‘epic’ is not indigenous to Homer. As has often been pointed out, the Homeric poems do not refer to themselves as ēpea, but rather as aoidē ‘song’ or kleos / klea ‘fame’ / ‘famous deeds’ (< kluo ‘to hear’): terms that obviously relate to oral performance. To be sure, Aristotle uses epos as a description of the Iliad and the Odyssey, and to this extent the term epic is, in historical terms, closely associated to the Homeric tradition, but it remains relevant, in fact urgent, to ask how much Aristotle actually understood about Homer.

Clearly, Homerists can profit from viewing Homer using a non-parochial understanding of the term epic. It also seems to me, however, that the classical local tradition has a contribution to make to the wider world of epic poetry. We have seen that Parry’s critics sometimes present themselves as trying to rescue Homer from the accusation of ‘unidimensional primitivism’. [49] Scholars working on African epic often present themselves in similar terms: rather than studying the preverbal myths of humanity, they are developing ways of talking about excellent performers and performances within a culture, about individual responses to epic traditions, about style, entertainment value, the rivalry and emulation between bards, and the responses of expert audiences. In this kind endeavour, it may well be that Virgil’s Homer has a role to play.