This excerpt is from the blurb on the jacket of my recent book:

‘The Dance of the Muses develops an authentic and at the same time revolutionary musical analysis of ancient Greek poetry ... A. P. David departs from the abstract metrical analyses of the past in that he conceives the rhythmic and harmonic elements of poetry as integral to the whole expression, and decisive in the interpretation of its meaning. Such an analysis is now possible because of a new theory of the Greek tonic accent ... and its application to Greek poetry understood as choreia ... David offers a thorough-going treatment of Homeric poetics: here some remarkable discoveries in the harmonic movement of epic verse, when combined with some neglected facts about the origin of the hexameter in a ‘dance of the Muses’, lead to essential new thinking about the genesis and the form of Homeric poetry.’

Sounds exciting, yes? A new theory of the historical Greek accent, and hence a new possibility of performing ancient Greek poetry ... a claim (not new, but always stimulating) that the epic hexameter began life as a dance—perhaps even the dance of the planets ...

There have now been three published reviews of my book, all of them negative, the most recent a perfunctory dismissal by Oxford’s M. L. West in the Journal of Hellenic Studies. For those who must defend their proprietary versions of Milman Parry’s oral theory of Homeric composition, the threat posed by my ideas is very real. In particular, for those who have spent a career teaching this oral theory as Homer 101—not as a controversial suggestion, but as a foundation for any serious reasoning and interpretation, in the manner of Einstein’s theory of gravity for modern cosmology—the embarrassment of misdirected authority, before students and peers across the academy (not just in Homeric studies or Classics), could well be acute. The instinct to distort, dismiss, and otherwise suppress my seventeen-year labour has a very practical origin for a senior academic.

But there are also of course those professionals who supervised, reviewed and recommended my work for publication at Oxford University Press, whom West tries to call to account. (‘OUP was badly advised in this case.’) The idea that dance was the origin of the Homeric hexameter seemed to some the most impressive. Most of them were sympathetic with some aspects of the oral-formulaic theory, but they all thought that my argument deserved a hearing nonetheless. That the book was published in 2006 is therefore, above all, a testament to academic freedom and indeed academic decency, at the highest level.

I did not in the least set out to challenge the oral theory of Homer, where it is supposed that generations of anonymous non-literate bards ‘grew’ the Iliad and the Odyssey, like twin olive trees from a single spot of earth, as they handed them down, memory to mouth to memory, in a performance tradition. I was raised on this vision, including in the writings of West. But the turn for me came in solving the problem of the Greek tonal accent: suddenly, unmistakably musical motives were revealed for the first time in Homeric verse, that had not been available or supposed in the entire modern history of Homeric scholarship. Only when it became possible to sing Homer, through a myriad variations and regular cadences, with effects spanning whole lines and passages, did the stitch-work of oral composition come to seem unsuited to the phenomena. It is the new and authentic way to perform all of ancient Greek and Latin, prose and verse, that is the lasting prize of the book, and not its stake in a modern Homeric Question. (Based on an analogy with the prosody of Vedic Sanskrit, which was first proposed by W. Sidney Allen, the theory applies illuminatingly to the classical Latin cognate as well as the Greek.)

The prevailing orthodoxy is well expressed by Gregory Nagy (Comparative Studies in Greek and Indic Meter, Harvard 1974): ‘traditional phraseology generated meter rather than vice versa’. (Although the thing is put in general terms, by ‘meter’ Nagy means only the dactylic hexameter; no other metre has been claimed to have been formed by traditional phrases; ‘feet’ are the more usual analytic constituents of a metre.) This claim has it that formular segments of Homer’s verse, of different lengths, were originally ‘lyric periods’, or cola, in some forgotten poetic and performance world. (The colon is a larger rhythmic grouping than a foot. Extant Greek lyric poetry, which supplies the length and form of the alleged prehistoric lyric cola, is demonstrably post-Homeric. This poetry, from Stesichorus to Pindar, exhibits a development over time from simplicity to complexity in rhythm—an historical development that does not require a prehistory.) Such segments are supposed to have been somehow amalgamated in this imagined prehistory into the full-length dactylic hexameter line. The characteristic word breaks of the Homeric line become therefore merely symptoms of surgery, unmusical birthmarks of this Frankenstein-like origin. Although there have been a number of competing theories about exactly which smaller rhythmic units may have been the building blocks of the mighty hexameter, no other metres or lines, inside Greek or otherwise, are ever adduced to support typologically the idea that their underlying rhythm has been stapled together in this way.

It is unclear how one gets from stapled beats from different lyric metres, to correspondingly stapled words and rhythmic phrases, that can end up meaning or sounding like anything, let alone tell a useful story. It has never been a standard which oral theorists have set for themselves, that they should be able actually to reproduce oral poetry. Indeed, the mysterious genesis of the hexameter itself, and then meaningful lines of hexameter, and then epic poems in hexameter, is shunted off into unknown realms and ‘dark ages’ of oral poetry, oral consciousness, and even oral palaces (or other imagined performance venues). There is a curiously noumenal appeal to the reconstruction of Nagy in particular, where the formative era of oral poetry is said to be forever lost to us behind the veil of the Homeric ‘transcripts’, ‘scripts’ and ‘scriptures’ that we have inherited (see Poetry as Performance: Homer and Beyond, Cambridge 1996).

The word ‘tradition’ gets bandied about, as indeed it must be in the interpretation of any conceivable work of art; but as though it had a special purchase in this case, because of the strange claim, unknown to any of the classical authors who inherited him, that Homer had actually inherited a lot of his phrases. Even without the claim, it is not illegitimate to speculate about a background to Homer, and passages in Homer. But there is in fact no evidence whatsoever, in music or theme, that Homer was traditional: quite the opposite. She was a maverick, from whom later artists in pottery painting, drama and philosophy took pains to diverge.

The romanticisation of oral cultures bespeaks a persistent sort of patronising colonialism in the Academy. ‘Strategies’ and ‘techniques’ in oral storytelling are confidently identified as prompts to poetic and cultural memory (as well as aides de mémoire), without any comparison to the identical techniques and forms in literate storytelling. Chiasmus—digressing and retracing one’s steps in reverse order (abc-cba)—occurs in the tale-telling and other sorts of accounting given by illiterate subjects right around the world. Chiastic rings (of a baroque complexity) also organise the argument of Plato’s Republic, however, and chiasmus is an almost ubiquitous feature of ancient rhetoric in poetry and prose. So how are we to locate the elaborate ‘ring composition’ that informs Homeric narrative? Is it primitive, or a mark of high culture? Perhaps it is like the chiasmus in baroque music, itself rooted in dance. My own answer is that it originates in epic when a storyteller digresses from a catalogue to expand on an element, retracing his steps in reverse to rejoin the list and proceed to the next item or event. (At an oral stage this expanding a catalogue like an accordion would better be called ‘intemporising’ than extemporising.) But it is absurd to suppose that Homer’s compositions contain a vast mnemonic key; they have to be memorised, just like other poems and songs. The right comparison for Homer is neither with oral storytelling nor with literature, for a very good reason.

The Muse is an Apollonian circle-dancer, and I of course believe that Homer’s dactylic hexameter is a dance rhythm.

It corresponds to a repeating segment of an ever-circling dance, a version of which is even today the national dance of Greece, the συρτός. From our earthly vantage, the planet-gods move eastward in the heavenly round with regular retrogradations; this is also the distinctive form of the συρτός. Its foot rhythm is long-short-short. Dactyls have always been dance feet, maladapted to speech. But even the small rhythmic groupings that were speech-like, were called ‘feet’ by the ancients. For me it was an epiphanic moment, at the very instant that it was a beer-can-crushed-on-the-head-for-the-sheer-stupidity moment, when I realised that ‘feet’ might mean feet. The steps of Greek choruses are set to these trochees and iambs, as well as dactyls and anapaests. But I suspect that even the iambic declamations of the tragic actor were choreographed in a way that might make Shakespeare’s seem like free verse.

The ‘choral theory’ of Homer rests on the assumption, justified historically and within the Homeric poems, that the dance came first. An hexametric segment of the συρτός is a whole thing, with internal articulations, not a sort of patchwork. This dance is mentioned in a 1st Century CE inscription from Boeotia, where it is already referred to as the ‘dance of the forefathers’. The characteristic word-divisions of the Homeric line correspond to moments in the rightward-moving dance, of turn, retrogression, and resumption. The retrogression is chiastic in its steps; the notion of ‘retracing one’s steps’ is literal in a distinctive segment of this movement. Those trained in a classical language will have heard of the two points of momentum shift, before and after the retrogression, as the trochaic caesura and the bucolic diaeresis. Hence Homeric phraseology, traditional or otherwise, in fact developed in response to an ambient, or circumambient, dance measure. (You may see a visual demonstration of the correspondence between Homeric verse patterns and actual dance movement here.)

It is certainly possible that my argument for the genesis of Homeric poetry is compatible with an as yet unformulated theory of oral composition, which would apply also to ‘rap’, a modern genre of chanted poetry which is extemporised to a repeating dance beat, and which uses samples of ‘traditional’ material. But once the concrete rhythms and turns of a dance enter the picture, it becomes pointless to maintain that the so-called ‘formulas’ in Homer are in fact pre-Homeric or even Indo-European building blocks of the hexameter line itself. What generative connection could there be between such verbal snippets and the larger, repeating metre of a still extant folk dance? It is difficult to imagine that a patchwork hexameter could ever have sounded like a whole line, especially if it is maintained that many of the verbal segments that make it up were actually whole snatches of independent songs in alien lyric rhythms—‘traditional formulas’—known as such to Homer and his audience through an oral inheritance. (‘Hey, Mum, why is that singer fella mixing up all our songs like that?’ ‘Shh, child, it’s just the new rap.’) Words don’t make dances. The dance comes first. But dances can be set to rhythmic groups of words which desire, like the dance, to repeat themselves.

Many features of the ‘lay’ experience of Homeric poetry, which remain dormant under the purview of oral theory, touch into life when they are supported by choral theory. The anaesthetic ‘traditional formulas’ are instead recurrent phrases—which is to say, musical recurrences. Anyone who has sung ‘verse’ and ‘chorus’ of a Christmas carol knows that she does not have to explain repetition in what was originally the accompaniment to a dance in the round: she would rather have to explain the lack of it. Sometimes these phrases were signature lines—‘swift foot Achilles’, ‘rose-fingers, the Dawn’—harbouring the musical power to evoke the presence of their objects.

In dance the evocation would have been an immediate revenance of ancestors and events in the midst of the ritual circle, as the names and phrases of the bardic catalogue were sung to the rhythm of the dancers’ feet. In the rhapsodic performance of Homeric epic, where an histrionic soloist stood before an audience who knew the traditional dance in its spine and in its toes, the evocation occurred in the narrative space between the rhythmic teller and his rhythmed listeners—much as it does, or did, in the ‘orchestras’ and opera houses of modern times.

Where a novelist controls, marks, and plots time, Homer ‘counts out’ the time. This is the sense of the so-called ‘singer’s rest’, in the arming of a warrior, in the laying out of a feast, or in the setting of the sun and the rising of the dawn. Such a method is obviously unavailable to the novelist, but it is practically forced on Homer, as he tells a story in the cultural and rhythmic ambience of a round dance that ‘counts out’ his very lines. The wonder is that he generally managed these distinctive interludes and transitions so artlessly. They are not singer’s rests at all. Neither are they refrains. They should best be called ‘listener’s rests’, which at the same time count out a narrative transition. When Odysseus is waiting for the sun to descend and the feast to end, so that he may at last get his escort home, the effect is sublimely metapoetic.

I am not offering an unproved assertion in the face of an established theory. The point is that both Parry’s argument and mine are arguments of the weakest kind: that is, arguments by comparison. According to Parry, repetitions of various kinds in the texts of Homeric epic are supposed to have been produced by the same forces that produce allegedly similar repetitions in the stultifying yarns of non-literate modern Bosnian guslars.

I suggest instead that repetitions occur in Homer most obviously because his text is musical: and one should as soon explain repetition in music as wetness in water.

But there is a rational way to judge between arguments by comparison. There was nothing deductive, for example, about the world-changing descriptive argument, first proposed by a magistrate of the Raj, that Sanskrit, Greek and Latin might be genetically related. Oral theory, by extreme contrast, does not answer to the phenomena: it requires each theorist to define exactly what he means by a formula; and these definitions still proliferate, after seventy-odd years. All reason and credit is lost when a theorist is led to stipulate ‘single use’ formulas. The theory itself was bankrupted when the facts of the case forced the major players to abandon Parry’s notion of ‘economy’—which was, after all, the logical underpinning of the whole idea. Stock words and phrases were thought to be drawn on by the extemporising singer to ‘fill up’ a measure—rather than chosen between alternatives, as in writing. Economy in his stock rhythmic phrases allows him to extemporise, without, as it were, thinking; a bit like a banker or an economist; but if there is more than one phrase to express the same meaning in the same rhythm, the chap has to choose. Hence without genuine economy in Homeric phraseology, there can be no oral theory of composition to distinguish Homer from a writer. (It is not generally recognised by the relevant scholars that ‘softening’ one’s stance on economy means also refuting the only essential premise of oral theory.)

Choral theory, by contrast, sings along with the dance. We answer why there are dactyls at all. We answer why Greek words had to be stretched, squeezed and coined by Homer. We answer why there are the unusual divisions in the movement of the line of verse. We answer why naming phrases, the ubiquitous nouns and epithets—what I call ‘choral signifiers’—are delimited in shape by these divisions in the summoning movement. We answer why the narrative tends to move in chiastic rings (the so-called ring composition). We even answer why so much of the narrative is about retrogressions within rings, in battlefield reversals and circuitous odysseys home. When there is a good comparison made, the phenomena answer back. Homer’s text is not like the classical music that emerges from dance, in the style of Parry’s claim that Homer’s text shows some signs of being like the humble products of extant oral poetry. Rather, Homer’s text is Music.

So sing along! And try the dance as well. For those who want to follow the debate at a specialist level, here are the reviews. First off is a clumsy hatchet job by Ronald J. J. Blankenborg in the Bryn Mawr Classical Review (2007.04.46), available here. My response in BMCR is here (2007.05.12).

Second was a review by Anne Mahoney in Versification (2008), which at least attempts to defend the various prevailing orthodoxies. My rejoinder is here.

Finally there is M. L. West’s review (JHS 128 [2008] 182-3), which is a brief attempt to dismiss and discredit the book. He cites only his own work in defense, and that without acknowledging it. I answer, almost interlineally, here.

On a less technical note, I append a lecture, ‘Some Home Truths About Homer’s Iliad’, which I gave in 2008 for the Basic Program of the University of Chicago. On this site as well (see the menu above) is my 2008 lecture for the Committee on Social Thought at the same university, ‘Epic Movement’.

A P David

Austin, Texas