Epic Movement

The Genesis and the Form of Homeric Poetry

A lecture delivered in the Committee on Social Thought at the University of Chicago, November 6, 2008

It is time for 21st Century Homerists to ask the following question in the most direct and unblinkered way: what exactly about Homeric poetry, in its broad and fine structure, is a theory about its oral, extemporaneous composition by anonymous bards, as part of an unkown and historically unattested tradition alleged to span uncharted centuries, supposed to be trying to explain? What exactly is meant to be illuminated about the extraordinary Homeric art works that we continue to read, study, and enjoy, by a comparison with stultifying orally extemporized yarns from Bosnia? I am not today going to refute the assertion of ‘economy’ that is, logically, the central pillar of Milman Parry’s output. All it has ever been is an assertion. I learnt the formal refutation twenty years ago from David Shive’s book, Naming Achilles, but the latest thing appears to be Rainer Friedrich’s Formular Economy In Homer: The Poetics of the Breaches. Friedrich is led to describe Homer’s text as ‘post-oral’. But without the economy of formulae that alone makes extemporizing possible, cognitively, there is no reason to connect the Homeric texts to any sort of extemporized origin. To judge by recent opinion, of every stripe, they no longer need an apology for their style. With far better justification would we call the Shakespeare Folio, in relation to historical performance, the composers who produced classical music in relation to dance, and the storytellers who published serialized novels in the 19th century, ‘post-oral’. I shall later advise you, based completely on internal evidence, post-what, in fact, the Homeric poems are.

As she manifests herself in ring composition within linear narrative; in the recurrence and the conjuring power of signature lines and epithet phrases; and in the distinctive sound of metrical diction: I hope to convince you that the soul of Homer’s poetry is born in dance.
This claim does not fit within the most recent modern orthodoxy, where you will hear many things about an ‘oral tradition’, and not a word about dance; prior to this orthodoxy there were hundreds of years of a so-called ‘Homeric Question’. I shall ask my own Homeric question this afternoon: how does a catalogue sung to the rhythm of a ritual circle dance turn into a linear narrative? It would seem that all the peculiarities and problems that create the field of ‘Homeric poetics’ can be solved in the investigation of this question. It turns out that ‘epic movement’ is a very real and graspable thing that not only links Homeric verse and narrative structure, but serves as a significant thematic source for epic narrative as well. The Iliad and the Odyssey are epic movements.

Most of you will have heard of Homeric ‘formulas’. Homer repeats a number of short phrases, and sometimes whole lines and series of lines, in the course of his compositions. Why? We should note that it is misleading to refer to these repeated phrases as ‘metrical building blocks’, as though we knew what was being built through the hexameter and why. The French alexandrine, for example, divides its twelve syllables right down the middle. Each hemistich is also divided, not always isosyllabically, but isochronously in performance. Hence there is a pendulous symmetry in its declamation. The epic line, however, almost never divides down the middle, through tens of thousands of verses in Homer and others. There is no imaginable linguistic reason that a central diaeresis should be prohibited in a dactylic hexameter. The notion that language generates metre in ancient epic—a central assumption for a number of prominent Homerists—runs foul on this and many other simple and obvious facts of Homeric usage. There must be a musical desire behind this rhythmic choice, based on notions of balance and a pleasing asymmetry, and of the line as a whole event, which determines the shapes of phrases whether they are repeated or not; only this desire can make any sense of the notion of a ‘formula’, as a thing that satisfies it.

First I address those theories that attempt to derive the epic hexameter itself out of smaller lyric units. The prevailing idea is that the allegedly formulaic subunits of an hexameter line began life as shorter lyric periods, which in some inspired prehistory managed to stitch themselves together into the long hexameter lines. One uses the reflexive voice because a poetic or rhythmic motivation for this stitching, and then its prolific imitation over tens of thousands of lines of epic, is not given. It is sometimes even suggested that the formula—not just the Greek language, but some of the actual rhythmic phraseology found in Homer—has an Indo-European heritage. Proponents of these ideas appear to think that they are consistent with the notion of the formula as metrical building-block; whereas the latter in fact assumes the metrical line as a given, in relation to which the formula becomes a compositional convenience. What they are truly proposing is the formula as a metre building-block—a candidate for Ripley’s, insomuch as it is a sort of thing that gives birth to its own parent.

There follow six intractable problems for the colometric derivation of the dactylic hexameter, with a seventh for the oralists who misguidedly champion this derivation. (I shall often be quoting from my book, The Dance of the Muses: Choral Theory And Ancient Greek Poetics, Oxford 2006.)

Problem the first for such a derivation is that no such thing has been demonstrated for any other stichic line, ancient or modern, Indo-European or otherwise. Poets (and musical composers) have always composed in whole lines, and groups of lines. Testimony for Homer’s aesthetic achievement in this regard is not lacking. People who respond to poetry with the sense that the poet is ‘filling up his line’, ought to have the courage simply to declare that they do not like this poetry, rather than explain it away (as, for example, ‘oral’), or become scholars of it. It could be argued that the English free verse of the last century was a poetry of fragments, after five centuries of pentameter rule. The centre did not hold. What we have, at least initially, is ‘line segments’ freed rhythmically from the context of the archetypal line. Even if this is not the true history of free verse, such a development is at least plausible. But to claim that the English pentameter was itself built out of such fragments—precisely the structure of the claim of colometrist ideologues in the face of the historical precedence in Greek of epic texts over lyric ones—is to try to put Humpty Dumpty together again.

Partial quotation in word, rhythm and melody, which is documented by [Gregory] Nagy in the case of Sappho and Homer, is not evidence of an integrity to the parts prior to the original, as is assumed in the concept of a traditional formula; partial quotation can be paralleled in other cases of musical revolution through history, whether we look to Renaissance polyphony in relation to plainsong, or to a bebop version of Cole Porter. Such quotation, nested within a new or merely idiosyncratic order, can help persuade an audience to get its bearings, and get hip. Snippets of hexameter phrasing in the mix give the verse legitimacy, by connecting the audience to its traditional music—at the very same time that the chorus seeks a new legitimacy in moving for the first time not so much to dactyls but to the rhythm of natural language. (David 241-2)

Nagy would like to claim that the curtailed Sapphic quotes are not quotes, but a direct tapping into the store of colometric formulae that is supposed to have predated the hexameter. I shall go on to address the fallacy (in the formal sense) involved here. But here is problem the first for a colometric derivation of the epic hexameter: there is not given any typological justification whatsoever for this kind of ‘derivation’ of a prolific, widespread, stichic line of verse.

It is a sign of desperation to claim 8th Century dates for lyric when no basis for such dating can be given. Here are the facts: lyric texts followed epic ones, and primitive lyric preceded complex lyric. The development in lyric is clear and historical and intuitive, for anyone with the judgement to distinguish Stesichorus from Pindar. But where did Homeric epic, the fully fledged chicken rather than the egg, come from? This is admittedly a mystery. But how does it help to put lyric first, in some proto-, ultra-primitive form with no conceivable exemplar—and make historical judgements that are not so much inverse, as literally perverse? Were the lost lyric forms complex and sophisticated enough to produce the panoply of epic rhythms and diction, only to revert to simplicity in the face of Homer, and recomplexify over time in the hands of Pindar and Sophocles? And why did these prehistoric lyric cola not agglutinate into other forms than the dactylic hexameter?

When it is understood that there is not really a competing suggestion, and that oralists have simply ignored the evident connection of the isometric hexameter to dance in their theorizing, perhaps my argument will be entertained at least with the suspension of incredulity which once greeted Parry’s theory, and did not immediately dismiss it as about so many monkeys at typewriters, generating an encyclopaedia over the extensible centuries of a ‘Dark Age’.

My criticism of both metrists and colometrists has to do with their falsifying mathematization of symbols that were intended to record rhythmic performance. In the book I address Nagy’s derivation of the hexameter from the pherecratean. The pherecratean is a catalectic or cadence metre in relation to a glyconic. 
One can hear the cadence effect well enough by ear, but if each syllable represented one step of the foot, the visual effect would perhaps have been even more emphatic. A glyconic that begins on the right foot would end on the left, and so a series of them can be strung together. The pherecratean, however, has one less syllable, and hence it begins and ends on the leading foot. In general, a catalectic metre is short one syllable in relation to its partner, and has an odd number of syllables; note that both of these properties make immediate sense if one moves to them physically.

It may at first sight make sense to say that a pherecratean has a synchronic relationship with a glyconic (i.e. it is the catalectic version) and that it has a diachronic relationship with the hexameter (it is the parent). The problem methodologically is that the nature of ‘metrical opposition’ is not analogous to phonemic or lexical oppositions in a synchronic state. To say that the glyconic/pherecratean relation is synchronic is like saying that 2 + 3 = 5 is a merely synchronic relation, or the relation between a fifth and an octave, or between male and female. The pherecratean does not exist without a glyconic; the catalectic version is derived from the original and exists in a definite musical relationship with it that is not determined by time or circumstance. There is not an opposition between them but a definite harmony. Every verse has to have a cadence: every glyconic or series of glyconics has to have a pherecretean. This is a fact of dance and rhythm, in no way analogous to a linguistic fact [unless something like ‘every verb has to have a subject’]. Think about ‘shave and a hair cut’ (bum bum ba bum bum), which has to be followed by—‘two bits’ (bum bum). In his diachronic claim, Nagy seems to think that you can isolate the ‘bum bum’ from its context—in which case it loses all its rhythm—and derive a whole new rhythm backwards from it, in anticipation of it as, once again, a cadence. But the pherecratean’s being as a cadence derives from its relation to its original partner; whereas the sequence of dactyls that Nagy grafts on to it would naturally rather seek a rhythmic cadence in cretic-based forms, to judge by lyric practice (as in dactylo-epitrite).  (David 163-4)

This brings us to problem the second, which is the problem with comparative metrics: it violates the most basic tenets of comparative historical reconstruction. It can define no isolable units, like the phoneme, which interact synchronically and maintain their identity through time. Comparative metrics, insofar as it presumes to wear the mantle and methodology of historical linguistics, is therefore completely bogus. Clever professors playing with signs does not a science make:

It is not illegitimate to isolate an element in a diachronic analysis—for example, the feature ‘voice’ in a consonantal sound change. The question becomes, however: what in fact constitutes an isolable metrical element? If one were to rephrase the question as ‘what is a rhythmical element?’ one would immediately see that rhythmic patterns and oppositions occur that are in no way susceptible to a synchronic/diachronic analysis. Rather, the contrasting elements of rhythm remain constant and universal as long as we remain bipedal creatures who draw breath, and as long as numbers retain their properties. (David 164)

But when we turn to one of the proposed examples, we turn from pseudoscience to simple nonsense. This would be the now iconic comparison of the eight-syllable Gayatri verse form to a Greek glyconic. Here is the comparison which fueled the colometric fantasy. The Sesame Street test fails: these two things

x x x x  c _ c _

x x _ c  c _ c _

are not like one another. This is problem the third. Even from a mathematical perspective, the invariant part of the glyconic has six elements, the other four. If you think there is nothing worth noticing in the difference between a half and three quarters, why not pay more tax. But it is the reading of metrical elements as individual quanta, rather than integral groupings of feet, that leads to a spurious comparison. A mathematician may be struck by an identity in the final four elements of the invariant portions, and perhaps you are as well. But what are the rhythms doing in each case? The gayatri always closes in iambs, an ascending cadence; whereas the glyconic always closes with a descending dactyl, immediately followed by a cadential cretic. One has to cut up this invariant dactyl to make the comparison in the first place. Hence the true comparison is between x x x x ia ia and x x da cr. As I said, the Sesame Street test fails.

But simple similarity is not necessarily the phenomenon that drives a reconstructive comparison. It took long years of exploitative militarism and commerce before a British magistrate in the Raj noticed something genuinely worth comparing between Sanskrit, Greek and Latin. So …

… What is the point that drives the comparison? It is claimed that they are both lines of eight syllables with a ‘variable opening and a fixed cadence’. Note that ‘cadence’ here has become a null term, meaning simply ‘ending’; to judge by native descriptions and usage, it was the pherecratean that produced a feeling of cadence, in relation to a series of glyconics that could be extended at the poet’s will. But more important to note is that this second feature connects the two patterns with virtually every known human rhythm. The English iambic tetrameter, for example, another eight-syllable verse, is also fond of trochaic variation in the first part of the line (think of ‘Déserts of vast eternity’, from Andrew Marvell’s To His Coy Mistress). Why should we pronounce any of these schemata ‘cognates’, and infer a common ancestor, other than the common rhythmic nature of our species as the generator of parallel forms? Disagreement finds its way to agreement; any human being knows this, who has had an experience of harmony or rhythmic satisfaction. It is not possible, or even prudent, to avoid sarcasm in this case. The proposed comparison and inference, and the generations of credence accorded it, discredit the whole project of comparative reconstruction. There is no rational reason to connect these two sets of symbols at all, let alone as a basis for comparative reconstruction of an eight-legged common ancestor of some kind. Problem the fourth.

(It is amusing to speculate about the reconstruction: does the dactyl in the Greek exemplum suggest the influence of a non-Indo-European neighbour? Or do we apply the rule that an unusual variant is likely to be original—a survivor in the face of the iambic IE juggernaut, which even took over Greek versifying?)

But we have not yet got to the salient point. In the Sanskrit line, the second and fourth syllables (in the ‘variable’ part) are usually long. In other words, the Sanskrit line is iambic, like Marvell’s tetrameter. The invariant part of the glyconic, however, always contains a dactyl abutted by a cretic (BUM bada bumba DUM).

This suggests that a concrete and distinctive dance step determines the form of the glyconic. It is not a bunch of stuff prior to an iambic punch. That invariant close in the Greek case is a dactyl abutted by a cretic, descending modulating to ascending. The gayatri verse ends in simple ascending iambic. The rhythms do not in the least resemble each other. They are as different as can be in the realm of the realities of rhythm and rhythmic expression … (David 240-1)

I draw on Antoine Meillet, who is generally championed by oralists, for his most important observation in this context. He points

… to the equality in length of arsis and thesis in the dactyl as ‘une innovation du grec’. [That is, the strong and weak parts of the foot have the same time length.] This fundamental isochrony in the foot, unique to Greek, is itself evidence of an orchestic origin for Greek metre. A language-derived metre would rather be expected to build itself out of contrasting time pulses, as Meillet well understood. An isochronous foot generates isometric music. Isometry is a prevalent characteristic of dance and of dance music. Neither Greek nor any other Indo-European language appears to have been designed to reinforce isochronous dactyls. (David 158)

The dactyl continues to be unique to modern Greek folk dance. The descent of Greek populations is admittedly controversial and its history politicized, but to deny a connection between the modern συρτός and ancient dactyl-based metres is to claim a sort of spontaneity in Greek soil and air, which mysteriously asserts itself upon its inhabitants, and makes them dance funny.

So apart from the fact that there is no rational basis for comparing the comparanda in order to reconstruct a common parent, the comparates, Greek glyconic and Sanskrit gayatri, are positively dissimilar in definitive ways. The invariant portions of the compared lines are not only of different lengths, but of unrelated rhythms. And in particular, the glyconic’s obligatory dactyl is unique to Greek. This is problem the fifth.

What we have next to deal with is in fact fallacy in the logical sense. Description is what it is: only extreme errors in historical description can lead to positive fallacy. A reviewer of my book speaks for the would-be orthodoxy:

It is also generally held that the forms of cola in Indo-European meter are derived from the formulae of oral poetics; this observation goes back to Parry and has been developed both within Greek and by comparative observation of other traditions (Lord, Foley, Schmitt). The recurring phrases of Greek epic have particular metrical forms, and the words and the rhythms grew up together. (Mahoney 1-2)

That words and rhythms grow up together in poetry is a pleasant and natural enough assumption, which may also be generally true. It is absurd, however, to apply this bromide to the peculiar case of ancient Greek, and to archaic epic in particular. Meillet’s observation that the dactyl was a Greek innovation is a distinction to be focused on rather than glossed. Pierre Chantraine provides voluminous evidence that the non-contrasting time pulses of the dactylic hexameter forced a violence on the language that could not be expected in a situation where ‘the words and the rhythms grew up together.’ As he puts it, without either fanfare or controversy: ‘il apparait que le rhythme naturel de la langue grecque s’adaptait mal à la métrique rigide de l’hexamètre dactylique.’ (see David 158-9) The claim, remember, is that phrases that were originally lyric cola became the formulas of epic hexameter. Here is the fallacy:

Epic hexameter phrasing everywhere exhibits phonological and morphological adaptations, necessitated by the metre, while lyric metres depend upon and conform to the native quantities of words: the ubiquitous metrical pressure on linguistic form observed by Chantraine in the hexameter is nowhere to be found in lyric pherecrateans. Consider the implications of this for Nagy’s hypothesis. Much of the ‘formulaic’ material in extant epic, which does display phonological and morphological alteration, does not belong to the period that actually produced the metre[, if ‘the words and the rhythms grew up together’]; it must rather be interpreted as a later product of assimilation to the hexameter, which, most paradoxically, had to have displaced such traditional material as did once generate the metre, and so did in fact fit the form euphonically and naturally without phonological alteration. How could such a displacement have occurred? This paradox ought to discredit any attempt to ‘derive’ the hexameter from smaller Aeolic lyric units (this is also [Martin] West’s approach). The paradox is that apparently ‘language-driven’ metres (Aeolics) are being asked to generate a metre whose extant poetry displays extravagant distortions of language. The maladaptation to Greek is not claimed for her lyric metres. (David 164-5)

The conclusion that follows from oralist premisses is actually a welcome one to me, and anyone else who finds Homer violently original and even satiric about tradition; but one ought not to accept conclusions from false premisses, and one very much doubts that oralists have thought these things through:

[Anyone] who work[s] from the premiss that the dactylic hexameter is a product of the Greek language, ought to consider how expensive is this assumption for [the oralist] belief in the traditional nature of Homeric diction. Once the facts of Chantraine’s description are taken into account, one is obliged to conclude that Homer’s language has displaced such native formulae as were required to generate the metre. Hence the premiss leads inexorably to the conclusion that Homer is non-traditional, that the extant diction of epic is an innovation, and that its meaning and implications can give us no direct evidence of either traditional language or thought as it bears on the Homeric poems themselves. An investigation into Homeric tradition through its diction must therefore begin by abandoning this premiss. (David 165)

This is problem the seventh, and it arises only for oralists, that Homer’s diction is non-traditional. Problem the sixth is the fallacy that language-driven metres like Aeolics can be used to generate a metre whose extant poetry ubiquitously displays extravagant distortions of language. The text of Homer cannot be composed of traditional formulas, if it was combinations of formulas well-adapted to the rhythm—lyric cola—that originally generated the hexameter.

In sum: the derivation of the Homeric hexameter out of smaller colometric ‘units’ fails because 1) there is no typological basis given for such a derivation of a line; 2) the study of comparative metrics, on analogy with comparative reconstruction in historical linguistics, is bogus; 3) the Sesame Street test fails when one compares Sanskrit eight-syllable forms with glyconics; 4) the basis for linking the comparanda (variation prior to invariance) is so broad as to link each of them to most known verse forms; 5) the rhythmic sense of the metres is unrelated and positively dissimilar, hence rendering implausible the possibility of a common parent; and 6) lyric cola, the supposed elemental constituents of the hexameter, do not in extant examples display the phonological and morphological adaptations characteristic of the Homeric text. The child is a monster! As an absurdist corollary, 7) Homeric diction, and therefore Homeric poetry, is definitively non-traditional.

*    *    *    *     *    *    *

Homer tells stories in rings. Why? The fine structure of Homer’s narrative pattern has been elaborately graphed by Cedric Whitman in his Homer and the Heroic Tradition. Here is one of his schemata, for Odyssey 8, circumscribing the Phaeacian dance performance that contains Demodocus’ song of Ares and Aphrodite.
There are many things to observe here, but note in particular the chiastic nature of this ring form; that is, that the way out from the central dance retraces the way in, in reverse order.

Is there a connection between the distinctive cadences of epic rhythm and formulas and rings, those celebrated but also notorious peculiarities of Homeric verse and narrative? Is there a larger concept that either encompasses or grounds these much studied phenomena, so that a more than arbitrary connection can be seen between the movement of a special kind of verse, and the movement of a special kind of speech and story that come to birth in its medium? Is there such a thing, after all, as ‘epic movement’?

Movement begins in rhythm: the dactyl that recurs in sixes. But such a recurring rhythm would not be a movement at all, if its endlessness were not somehow articulated. The hexameter’s rhythmic ‘personality’ must be the ground of anything one might wish to call ‘epic movement’. The opening line of the Odyssey well illustrates the two most basic articulations of the epic line:

νδρα μοι ἔννεπε, Μοῦσα, πολύτροπον, ὃς μάλα πολλὰ

These are the trochaic caesura, after Μοῦσα, that divides the line centrally but asymmetrically, and the bucolic diaeresis, between πολύτροπον and ς μάλα πολλὰ. What is the origin of these ‘cuts’ (caesurae within the dactyls) and ‘divisions’ (diaereses between dactyls) in Homeric poetry? A commendable instinct in linguists suggests that the answers to metrical questions should be found in the structure of language itself. But the case of Homer presents some obvious obstacles to this view. Chantraine gives numerous examples of phonological and even morphological adaptations, various ways in which Greek must be stretched, squeezed and coined to fit metrical constraints in Homer. Hence one must look for an extra-linguistic source for Homeric metre, that produces these constraints.

Diaeresis, the conjunction of a new word and a new foot, marks the beginning of a line—there are no hyphens in epic—and hence it constitutes an inceptive cue. How can one account for a regular new beginning of movement, the bucolic diaeresis, immediately before the end of the hexameter line? The standard introductions do not seem to feel even the pressure of a problem here: accounts of the diaeresis as a so-called ‘rhythmic clausula’ do not pretend to address the question of why such a strange effect might have been a poetic desideratum.

Oralists and metricians generally have attempted to analyse Greek poetry without a theory of where to stress the words. Metrical study has not proceeded in this way in the case of any other known language, including Latin. It has been assumed without argument that word-endings must have been the most salient prosodic feature in this poetry, despite the knowledge that Greek did not automatically accent either a final or an initial syllable. It can only be the length of these purely academic traditions that now blinds us to the implausibility of such a method and its assumptions.

A new theory of Greek prosody that I developed in Chicago has now been published in my book. I there show how W. Sidney Allen’s discovery of an apparent stress pattern in ancient Greek, derived from the study of line ends in ancient poetry, was connected to the received system of accent marks introduced by Aristophanes of Byzantium. The key to the new theory also came from an observation of Allen’s: that there must have been in Greek, as in the cognate Sanskrit, a svarita, or automatic down-glide following the rise in the voice that was marked by Aristophanes’ acutes and circumflexes. Hence the Greek accent was a ‘contonation’, with both a rising and a falling element. Where this down-glide occurred on a long vowel or heavy syllable—not just in the case of vowels marked with the circumflex, but also in the case of long syllables that immediately followed the acute mark, syllables that were themselves unmarked—my theory argues that it was these long or heavy syllables, ‘barytones’, that were in fact the most prominently stressed elements in a word.

Here is a rhapsodic version of the invocation to the Catalogue of Ships, done according to the new theory:

σπετε νῦν μοι, Μοῦσαι Ὀλύμπια δώματ᾽ ἔχουσαι

μεῖς γὰρ θεαί ἐστε, πάρεστέ τε, ἴστέ τε πάντα,

μεῖς δὲ κλέος οἶον ἀκούομεν οὐδέ τι ἴδμεν

οἵ τινες ἡγεμόνες Δαναῶν καὶ κοίρανοι ἦσαν.

πληθὺν δ᾽ οὐκ ἂν ἐγὼ μυθήσομαι οὐδ᾽ ὀνομήνω,

οὐδ᾽ εἴ μοι δέκα μὲν γλῶσσαι, δέκα δὲ στόματ᾽ εἶεν,

φωνὴ δ᾽ ἄρρηκτος, χάλκεον δέ μοι ἦτορ ἐνείη,

εἰ μὴ Ὀλυμπιάδες Μοῦσαι, Διὸς αἰγιόχοιο

θυγατέρες, μνησαίαθ᾽ ὅσοι ὑπὸ Ἴλιον ἦλθον ·      (Iliad II. 484-92)

We may begin to consider the cuts, divisions and other terminations characteristic of the hexameter as cadence points for the prosody of Homeric song. Look now at the metrical charts for the opening lines of the Iliad and the Odyssey:

On the left, overlaid upon the actual syllabic quantities, which vary in the weak part of the foot from line to line, are the positions where the new theory locates the accentually prominent syllables. (I have used an acute sign to mark oxytonic prominence, a grave to mark a prominent barytone, according to the new parlance.) On the right we see the locations of the written accent marks (I have used simple acutes). In the new theory it is assumed that each word that is not augmented by an enclitic, regardless of its whole melodic profile, has a single most prominent prosodic moment—either oxytone or barytone, according to a definite rule. It is also assumed that, with the exception of the final foot, the ictus or metrical beat always occurs on the initial long, the thesis of the foot in this descending rhythm. Hence the charts reveal almost immediately the places of agreement and disagreement between accent and ictus.

A once hidden pattern has been disclosed. The rhythmic-harmonic analysis of passages from Homer presented in my book shows clear evidence of a tendency towards agreement not just at the end of the line, where one should expect it in any case, but also in the third foot. The sign of this is accentual prominence on the long thesis of the third foot. The written marks do not show this agreement, even at line end, as you can see; the new theory reveals it for the first time. (Note that in the right hand charts the written accents, corresponding to the modern Greek stress positions, occur more often than not in the arsis, or weak part of the foot. This is the anomaly that has heretofore made the tonal performance of Greek poetic rhythms an intractable problem.) Caesura can now therefore be seen as an automatic consequence of the desire to accent the thesis of the third foot, to produce a mid-line cadence. An asymmetric mid-line cadence can only be understood musically from the perspective of the whole line; obviously it has no meaning from the perspective of the colon or clausula. Caesura in Homer results from the prosodic placements that produce a sense of agreement—a concrete musical motivation—not from a need to pour words into traditional metrical moulds. The Greek recessive rules allow for only two possible locations of barytonic (heavy) stress: on a long penult (followed by a short) and on a long ultima. Hence barytonic prominence placed on the third thesis of the hexameter, producing either a masculine or a feminine cadence at that point of the line, entails either the penthemimeral or the trochaic caesura. That is why there are the two of them. That is why these word breaks exist in the hexameter.

The Ancients themselves have bequeathed us the notion that the elements of a metrical pattern are ‘feet’. Ancient Greek verse was either danced or danceable. Plato in the Philebus is explicit that rhythms and metres are measures not of speech but of bodily motion (17c-d). The entire list of descriptive terms generated by poets and grammarians, such as arsis, thesis, the foot itself, period, strophe and chorus—not to mention the use of the verb βαίνειν and the noun βάσις to describe the performance of a foot—must either be ignored if one denies the connection to dance, or be understood to participate in some delusional metaphor of societal and cultural compass.

Within the Odyssey itself there are a number of depictions of dance accompanying epic song, but singers are also depicted as soloists. The emergence of the hexameter as a purely sung or recited phenomenon may be compared to the submergence of dance rhythms in the modern classical repertoire, where soloists, the so-called ‘orchestra’, and the audience alike remain seated. When Alcinous wants to impress his enigmatic guest with the skill of the Phaeacians’ dancing, however, after they have embarrassed him somewhat with their showing against Odysseus in athletics, the community reverts to a full, epic display: the nine appointed officials smooth out the orchestral space, Demodocus proceeds to the centre, and a group of bachelors, in their first maturity and said to be skilled (δαήμονες) at the steps, circle him and beat the divine dance ground with their feet. They are therefore not casual improvisers, responding to the singer’s rhythm; rather, the dance comes first in this depiction, before the song begins, and appears to continue through the song. The dance comes first. Surely this is why Greek phonology has to be adapted to quantitative dactyls, if the singer is to sing along and syncopate in Greek.

There is a modern remnant of ancient dance that survives in the folk tradition of Greece, the dactylic round dance I mentioned called the συρτός. It was performed as part of the closing ceremony at the 2004 Olympic Games in Athens. The name carries the stamp of antiquity; a Boeotian inscription from the 1st century CE refers to the dance of the συρτοί. While this is a late date in relation to Homeric or classical times, the dance has apparently survived for nearly two millennia since then, and what is more, it is referred to, even in the first century inscription, as the πάτριος ὄρχησις: the dance of the forefathers. (Taken as descriptive rather than limiting, the adjective πάτριος yields an even more intriguing sense: the dance of the ancestry, that is, the catalogue dance.) This dactylic round dance of the twenty-first century clearly has a prodigious history. It has been neglected by philology as a clue to epic movement and form: what is remarkable is that in its articulations we find those originals of the epic movement, the bucolic diaeresis and the trochaic caesura:

Here we find an answer to our puzzle about the diaeresis. A resumption of rightward movement after the retrogression in the dance corresponds to the ‘second beginning’ in the words. Hence dance is likely the source of this peculiar feature. Hence in addition, the division in the συρτός suggests a reason for a favouring of the trochaic over the penthemimeral caesura. Although both appear in Homer, the trochaic is evidently the preferred form. In the modern folk dance the locations of these classical divisions of the line frame a distinctive retrogression and then a resumption in the circling step. They mark the ‘tropic’ points of a dance that revolves with involutions. It is perhaps not a coincidence that the word πολύτροπον (‘of many a turn’) exactly fills in the retrogression between caesura and diaeresis in the opening line of the Odyssey. Not only the man, but the verse itself turns here.

Now look at a video. The dancers were liberal arts students at St. John’s College, Annapolis, who prepared in their spare time and had no previous experience of Greek dance. You will recognize the invocation to the Catalogue of Ships. No claims of authentic performance practice are here made. Note only how the articulations of the συρτός correspond to the caesura and diaeresis of Homer’s verses.

Please watch at least the first minute or two.

I should point out that in dancing as a group to the ‘speech of the Muse’ (λέξις Μούσης), we not only imitated the Phaeacians, but followed a specific prescription from Plato’s Laws (795d-e) for gymnastic education. But I should also point out that as far as I know, there is no record of Homer’s poetry actually being danced to, before February 2001, at the first of these workshops. By the same token, there is no evidence that the opening chorus of Bach’s St. Matthew Passion has ever been danced to. All the same, it is a siciliano, and the music moves, the music itself dances, in a directly recognizable way. In a similar sense, all of Homer’s music is dance music. Just because Bach may have composed seated at a keyboard does not mean that the rhythms and movements of dance were not animating his fingers; the same may be said of Homeric composition and its composers.

What is a metrical building block? It is a segment of dance. What does it mean to ‘dance out’ the name of an object, flush with its native epithets? It is to conjure the presence of the referent, whether human or otherwise, in the locus of the ring, through the choral expression of movement and song.

The earliest attested hexameter poetry is epigrammatic, but for many reasons scholars consider catalogue poetry, and the catalogue portion of Homer’s works, to be the deepest historical stratum still visible on the surface of epic. Catalogue poetry is surely, from a literary point of view, the most boring portion of the Greeks’ poetic legacy. But consider what the effect of this poetry must have been like in a danced performance. What begins as a rhythmic, regularly retrogressing movement in the round, takes on a semantic force as the song strikes up, and the rhythm and harmony of the ancestral names interweave with and, as it were, re-harmonize the rhythm and direction of the steps. And conversely, just as the dance becomes meaningful, so also does the word in dance take on the power of circle magic, so that it not only points, but summons. As one danced to the florid chant of names in their rhythmic ideality, one felt the very presence of one’s ancestors gracing the communal circle: the storied warriors and their well-balanced ships on the expedition to Troy, or the noble women of the past in the matriarchal line. The performing of a catalogue was not therefore a history lesson, or a mere exercise in memory: perhaps it was in fact something like a public séance.

There is no intrinsic or necessary connection between catalogues and dance. In other words, catalogues need not be metrical. Τhe archetype of a catalogue is the series of counting numbers, a list of proper names in a fixed and unchangeable order. That is, one remembers that one element follows another in a catalogue or list (whether shopping or genealogical) in the way that one remembers that six follows five; and the way that one remembers the latter is lost in the very first functioning of the active memory. To change the order is unthinkable: the elder who recounts the village ancestors must begin at the beginning of the begats—he may not start in the middle and is unable to ‘zoom’ right to a particular. We have all experienced the profound disruption involved when one is interrupted while counting something.

Dance adds to the speaking of a catalogue, which is a list of substantives, the phenomenological summoning of each substantive; and in addition, it provides a rhythmic template that delimits the expansion in proper names and epithets of a danced-out signifier. What natural limit could there be, after all, to the proper names and titles of a god whose living presence one wished to entice? The internal articulations of the hexameter that we have derived from dance provide the delimiting template for expanded naming. Hence the combination of catalogue and round dance provide a complete account of the genesis and the ontology of metrical name-and-epithet phrases in Greek epic. The idea here is quite simple: that to name something in a circle dance, so as to tap into the dance’s summoning power, one must name it in the rhythm of the dance, and according to its breaks or changes in direction. Thus the phenomenon of caesura and diaeresis as name and phrase boundaries in Homer.

One does not need to explain the recurrence of words and melodies in the accompaniment of a round (such as a Christmas carol): one would rather have to explain the lack of it. While repetition in a written text is profoundly meaningless, almost the very type of meaninglessness, singers and composers (and orators, in their fashion) have always understood the musical cogency of echoing and repetition; that in the semantic purity of music, repetition is the principal way to create context and meaning. Living speech is a musical phenomenon: and one would as soon explain repetition in music, as wetness in water.

The power of such phrases survives, in the transition from bardic round dance to solo rhapsodic declamation, because its rhythmic pulse still governs a noun-and-epithet phrase, and so propels and infuses its capacity to name. The histrionic dimension of text and performance continues to be rooted consciously in the recurrent cycle of dancing feet, such that Aristotle in his time can number and divide the steps of an epic line with syllables:

τὸ ἔπος δεκαεπτά ... βαίνεται δ᾽ ἐν μὲν τῷ δεξιῷ ἐννέα συλλαβαῖς, ἐν δὲ τῷ ἀριστερῷ ὀκτώ.(Metaphysics 1093a29-b1)

the epos is seventeen … and it is stepped on the right with nine syllables, and on the left with eight.

Note that Aristotle speaks of the syllable as a thing that one steps.
This usage, where a syllable is a measure of a dance, is a welcome and long-missing complement to the notion of a ‘foot’ as a measure of poetic speech.

  Consider μος δ᾽ ἠριγένεια φάνη, ῥοδοδάκτυλος Ἠώς, the Odyssey’s characteristic evocation of daybreak: ‘when she, the early-born one appeared, rose-fingers, the Dawn.’ This verse and other Homeric recurrences have been well described as ‘brief incantations’. The epithets become names, transporting a single feminine subject—hidden in the Greek until she is revealed in the proper name of Dawn—through the line. The intoning of this line is a dancing out of its subject that culminates in an evocation and an embodiment. As the line comes to rest in the balanced modulation from descending to ascending rhythm in οδοδάκτυλος Ἠώς, we feel forever the spread and fall of morning rays, and their consummation in the emergent reality of risen dawn.

For all that the very meaninglessness of repetition dulls the semantic power of a phrase, it must be stressed that rhythmic repetition is the only way to draw out a phrase’s musical quality; let me repeat that, the only way; the prize is what could be called a choral signifier, a word fully realized in its evocative, musical potency.
Here is the solid empirical reason for repetition in the literary text of Homer. As profound a receptor of Homeric poetry as Plato understood that there was a λέξις Μούσης, a ‘speech style’ of the Muses (see Laws 795d-e). Hence the repetitions and the formulas, with which classical music is also full, were received by Plato as features of a certain dancing style characteristic of the most exquisite compositions—not in any sense as the hallmarks of an extemporized origin.  

I would maintain that it is possible to infer that the text of Homer constitutes the end-product of an aesthetic development, purely from evidence within the text of Homer—the Spinozist standard; but also that it makes absolutely no sense to refer to the result as literary, on the one hand, or oral on the other. Neither is it ‘post-oral’. Orality only enters the equation on the level of the list, a pre-Homeric level, a handed-down sequence of people or events, which may or may not have been expressed metrically. The Homeric text that we have is actually ‘post-catalogic circle dance’.

The expansion of the original catalogue-dance or catalogue-poem by means of loops ‘picked up’ by relative clauses, ‘she who’ or ‘they who,’ was a simple and likely spontaneous response of the cataloguer becoming a tale-teller. Oralist theorizers also note this expandability by means of relative clauses as a contributor to the technique of extemporizing. What they have missed is the function and gestalt of the frame for these expansions. The motive of extemporizing is deceptively similar but in its root notion anathema to that of an ‘intemporizing’ catalogue poet: here instead we find a notion of preserving a memory in an absolutely distinct and definite order (like that of a genealogy), while expanding within it in ways that must not disrupt that order. A story-teller caught up in the transports of the moment may well forget his place. Hence when he expands on some or each of the terms in a list by evoking its life-giving elements, he then must take special care to rejoin the list. Often he makes the digression circle back on itself chiastically with a repetition of themes in reverse, initially perhaps as a mnemonic aid, before proceeding to the next term. He thereby both remembers and preserves the relative temporal order of the catalogue: he is able to digress as expansively as he desires, but returns via the path of his introduced themes in reverse order, so that his digression becomes a retrogression. Thus both the mechanism and the aesthetic of ring composition spring from one source. By means of it this sort of tale-teller always does his imaginative and expansive poetic work, but also always keeps his place, within the larger frame of the catalogue.

A ring form need not be chiastic. An antistrophe in lyric, for example, is a backward circling, but its rhythm recapitulates, rather than mirrors, the rhythm of the strophe. Hence chiastic order is a specifically retrogressive and epic form of ring composition that points in sympathy to the origin of all epic movement in a peculiar, dactylic dance, characterized by a pivotal retrogression. In such epic movement there is a way forward and a way back (abc-cba) that delimits a retrogression within an onward cycle. Hence there is a meeting of motives here—need transforms into desire: a cataloguer’s need to re-find his place matches the aesthetic sense of a peculiar movement.

An epic catalogue is like an unexpanded accordion. Within the fixed sequence of Odysseus’ wanderings, the Lotus-Eaters receive only a few choice lines; but Odysseus expands the Cyclops episode with all his Homeric resourcefulness in narrative, similes and speeches. A necessary lop-sidedness that results from the unequal treatment of epic ‘intemporizing,’ as compared with the symmetries and climaxes under the novelist’s control, is balanced somewhat—in the case of the Odyssey at least—by a sense that the teller of the Wanderings has imbued and concentrated each segment of the story with the whole of his distinctive vision. That is to say, even in the twenty lines of the story of the Lotus-Eaters, he has distilled the gist, of νόστος and forgetting; so that the episodes are like differentiated cells, each carrying the whole set of the organism’s chromosomes. (Note that, by contrast, in the ‘catalogue version’ of the Tale, told to Penelope in their bedroom, the Cyclops episode gets only two lines—less there than either Aeolus or the Laestrygonians.) This thematic unity is perhaps an indicator that it is one poetic consciousness that is both traversing and expanding on the elements in the telling of the Tale. But, clearly, the power in such meta-sequencing of received catalogues can only hint at the power wielded by the novelist over time, in his created narrative universe.

Note that we also find the relative clause as an entry into mythic digression throughout the extravagantly idiosyncratic poetic forms of Pindar; Pindar’s ‘intemporizing’ rings attempt something truly daring in relation to epic, in that the myths are often told in reverse chronology. That is, backwards. Hence ‘epic techniques’, which can be confusing to a modern or otherwise foreign sensibility, came to be aesthetically choice worthy in later generations and later poetic media. There is no reason to connect them to extemporizing bards. Rather, they should be seen as techniques growing out of an aesthetic project that is common to both epic and lyric in Greek: the choral unfolding of linear myth.

What is possible here in the ‘in between’ spaces, the suspensions in a catalogic movement, is μῦθος, that interesting word for ‘word’ in Homer. A conspectus of Homeric usage allows us to infer a sense for it of  ‘disclosure’. The framework of the plot recedes behind moments of intensity and revelation. The framework cannot disappear, it must be returned to, or else the order of the whole, not in this case a cosmos but a sum, would disappear. Yet such moments define centres and are the first propagants towards true cosmos in the Homeric narrative. Again, these occur not in the joints of the frame, but in moments in between that must be prepared for and receded from.

The most obvious examples of such μῦθος, indeed the titular exemplars, are the speeches. Each is a disclosure of what is hidden behind the mask of name, face and status. Τhe narrator puts on that very mask and expresses what is indwelling at the various loci of individuality in the human figure, as though he were in fact seated at the seats of such a person’s consciousness. Homer is not far different from a dramaturge, in that a tragedy is formally a series of entrances, stands, dances and exits of a chorus, but the moments that bring unity to the whole are moments in between of concentrated rhythmic eloquence in the actor of a great figure, who has lesser figures and lesser speeches for foil. In Homer of course the speeches have a Shakespearean realization and potency. Little in the world’s dramatic literature can compare with the multi-part tour de force of Iliad I. But even the first attempt of the catalogue-singer to reproduce a divine or human speech in epic metre had to have been, in the nature of things, a fully-formed entrance into the world of art and the representation of thought, like the birth of Athena from the head of Zeus.
One need only present a Trojan deliberating, whether to stay by the oak, or to go out into the plain to confront a raging Achilles, to immerse an audience fully in the notions of ‘choice’ and ‘dilemma’ that characterize human experience.

One had always to remember the frame, however. A speech in a novel is in some way integral to the plot, which is a larger, authorial, and climactic structure. A speech inside catalogue poetry, on the other hand, is like a passing window into life and form that yields merely to what or who must come next.

Most remarkable in the Odyssey are lyric moments that are not speeches, but what could be called narrative μῦθος. One such is the description of Phorcys’ cove at which Odysseus arrives in Ithaca, with its olive tree and its sacred cave with two entrances (13.96-112). A favourite is the description of the twin bushes at Odysseus’ arrival in Scheria (5.476-93). There is an opportunism in this imagery, a kind of reaching for wholeness from within the interstices of a simple recounting. (Note how ‘re-counting’ directly translates the concept in καταλέγειν, whence ‘catalogue’.) The entrance of Athena at the end of the image of the intertwined trees and fallen leaves, to pour sweet sleep upon the eyes of Odysseus, is essentially and mysteriously—that is, revealingly—authorial. A consciousness peers out through the trees.

We speak of ‘suspensions’, ‘framings of intensity’, of ‘lyric stasis’ as intrusions into an onward movement that is characteristic of Homeric epic. The similes as well are an excellent example of this. Perhaps it is possible to see the original of epic movement in the motions of the planets. Even the dance is not the original: it is the planet-gods who, from the vantage of the earth, circle ever onwards to the east, but with regular westward retrogressions apparently tied to the sun. As an outer planet approaches opposition to the sun, so that it sets as the dawn rises, it slows down and appears not to move against the starry background for some days. In ancient astronomy, this point is referred to as a ‘station’. It then retrogrades through the opposition until the planet comes to a second station, appears again to remain perfectly still, and then seems to resume its course—retracing its steps to complete a contemplative heliacal interlude. Hence the planets are not only polytropic: their apparent motion is characterized by stasis and chiastic reversal as they come each time to face the sun. Hence they exhibit what we may now recognize as ‘epic movement’.
Even the plot of the Iliad, taken as a whole, can be understood as a retrogression. The father of gods and men spends a sleepless night at the beginning of Book II, trying to figure out how he can do the will of Thetis—her ξαίσιον ἀρὴν, a prayer outside of fate (XV.598). The problem involves stitching in a sequence of events inside a grander schema that has already been determined. This amounts to a problem of historical—for Zeus—and hence narrative—for Homer—retrogression. Ilium is bound to fall; but for a time there must be a reversal: the Achaeans must suffer, ten thousand griefs, because of Achilles’ vengeful wrath. The tide of battle sweeps backwards from its goal, from Ilium back toward the ships, and then flows forward again to arrive at the death of Hector beneath the walls. One does not end up where one started, however, because time itself never stops or reverses, and the turning points are marked by deaths. In Book VIII we learn from Zeus—perhaps he learns it then too—that Patroclus is to die. In Book XV, once he has awoken from the slumbers of spent love,
Zeus reveals for the first time that Hector also must die, and so the story’s horizon comes into view. (Zeus there announces the reversal in battle, back from the ships to Ilium, with the word παλίωξις, ‘retrogression’, in the post-caesura position.)

The interplay between recited catalogue and disclosive art is close to the surface for this poet. One sees this best where Homer stretches, as when Odysseus seems cruelly to manipulate his father with mocking disguise; but the payoff is the moment of disclosure that calls forth a memory which would have been far less potent without such an introduction. Odysseus recounts a gift from Laertes who named and numbered the fruit trees and vines as Odysseus trailed after him, a boy in the orchard (24.336-44). Odysseus remembers the numbers, thirteen pear and ten apple trees, forty figs and fifty vines: a catalogue that anchors him to a time, a place, and a father. In this case the catalogue is the disclosure, and we ourselves are justified in recognising a kind of metapoetic inversion behind the episode’s pathos. The Homeric task of bringing a catalogue to life is here achieved by the sourcing of a life in a catalogue.

In the tension between the remembering through catalogues and merely temporal, discrete sequences, on the one hand, and the articulating of affect and of meaning through the enactments of speech, simile and narrative epiphany, on the other, there is created a most fertile seedbed for poet and critic both.
Twin trees spring from one spot of earth. The more and more intricate branchings of the analysts, who take Homer apart at the joints, are met twist for twist by the insinuating tendrils of the Unitarians, who keep experiencing an organism in between. That there is a poetic necessity answering to the critical schism: this is perhaps a new way to comprehend the phenomena of Homer’s modern reception.

There is an ethos to the helical narrative style that is revealed in a rather self-conscious metapoetic parable. Elpenor tells Odysseus in Hades that when he was laid out to sleep drunk on Circe’s roof, he failed to pay attention and go back down along the long staircase, but instead fell straight through the roof; his neck was separated from his vertebrae and the soul went down to Hades (11.62-5). The fault is not that he presumed to climb so high, drunk as a god, and to take his rest on the roof; the fault is that he made his way back straight through, and failed to slant (or circle) down the stairs the long way, returning the way he came. Odysseus of course is not permitted to make his return straight either; he must wander, circle and rest a long long while. The deft part here is that Elpenor himself is the mechanism by which Odysseus is forced to make a chiastic circle back from Hades to Circe’s isle before he can continue his journey home. The journey to Hades becomes, by means of Elpenor, the centre of both a physical and a narrative retrogression. In life and in story-telling, as in dance, there is a correct way of returning by turning. An ethos of the danced hexameter becomes an ethic of the narrative, in one epic movement. The head stays united to the spine.

Somewhat unexpectedly, Penelope looks back at Helen in the midst of the very speech in which she recognises and embraces Odysseus, back to the woman at the heart of it all—not perhaps of the story of Achilles, but the cause of the whole cycle, an energy at the periphery spinning the orb like Aristotle’s prime mover. She begins by excusing her reticence, explains how she had learned to be afraid of deceitful men; but then seeks a most paradoxical topos in Helen, that serves subtly to shift the focus to Odysseus’ mistakes, and to provide us the single most synoptic view of both epics from inside the poems.
Not even Argive Helen, she says, would have slept with and enjoyed a foreigner, if she had known that the warrior sons of the Achaeans were only to bring her back home (23.218-21). Would Odysseus also have left if he had known he was simply to return? She has come to see the whole in a circle that circumscribes both her husband’s and Helen’s folly, a coming full circle in the pattern of fate that encompasses the impetuous linearity of a woman’s desire and the futility of a man’s endeavour.  From this perspective—transfixed by the sudden, aged awareness that there has been an immense retrogression—perhaps there can be reconciliation, and the resigning oneself to it.

What is left for an oral theory? The mnemonics of catalogues are indeed an oral phenomenon, but they do not necessarily yield us metrical poetry. What is at least conceivable, however, is that a number of popular formulas may have developed and circulated among bards who attempted to sing along to a dactylic round. (The six-foot grouping itself may have been an imposition of the accompaniment upon an eternally circling dance movement.) The endless variety of blues improvisation is drawn, after all, from a few simple clichés. These formulas may have become ensconced in the vocabulary of hexameter story-tellers, who came to compose and perform neither as, nor with, accompaniment. But those who believe, inversely and perversely, that verbal formulas generated the hexameter itself, can never accept that the extant epic poetry of Homer and others was connected to folk dance, or even that depictions of danced performance within Homer are depictions of the hexameter. Quite apart from the fact that the phrases of lyric cola would not show phonological distortion, or the problem that there is no plausible historical connection between lyric cola and any sort of extemporized verse, it is simply impossible that such verbal phrases somehow created the repeating, circling rhythm of an archaic folk dance. The dance came first. And, I would suggest, the way that ghostly catalogues can take on flesh, and haunt the present in performance, was an intoxicating inspiration for the narrative artist who produced the sublime Iliad and the sad Odyssey, poems that we now read recumbent, reveling yet in wine dark seas and swift-foot Achilles.

‘Epic movement’ in all its levels, dimensions and sweep, is only possible because the human foot sometime diverted from its purposes backward and forward, to bear its burden sideways with bended knee, and a twist in the back, so to circle (with retrogressions) in the way of the cosmos and the dance of the Muses. There is no level of Homeric art that is not informed by the chiastic aesthetics of retrogression. My own sense is that Homeric scholarship, whether oralist or scripsist, is trending toward the greater recognition of structure, complexity and self-referentiality. (No such development has emerged in the study of Bosnian epic.) Homer’s poems no longer need an apology for their style. A number of us, a critical mass, have been humbled by them, and long to speak this experience. Perhaps we have come so far from the spring, that we have forgotten to thirst—or we are being watered by secret springs. Oral theory was intended, after all, to account, divertingly and creatively, by 20th century standards, for a certain primitiveness of style perceived by continental-bourgeois critics in relation to literary epic, and later forms of narrative. In a new, pre-oral world, have we at last forgotten this? Shall we not drink from the waters of Lethe, and clasp hands—and dance?


Allen, W. Sidney, Accent and Rhythm. Cambridge: Cambridge University Press, 1973.

———    Vox Graeca, 3rd edn. Cambridge: Cambridge University Press, 1987.

Chantraine, Pierre, Grammaire Homérique. Paris: Librairie C. Klincksieck, 1942

David, A. P., The Dance of the Muses: Choral Theory and Ancient Greek Poetics. Oxford: Oxford University Press, 2006.

Fitzgerald, Robert, ‘Postscript’, in Homer: The Odyssey, tr. Robert Fitzgerald. New York: Vintage Classics (orig. 1963, Doubleday & Co.), 1990.

Friedrich, Rainer, Formular Economy in Homer: The Poetics of the Breaches. Stuttgart:  Franz Steiner Verlag, 2007.

Mahoney, Anne, ‘Review of The Dance of the Muses: Choral Theory and Ancient Greek Poetics by A. P. David’. Versification (arsversificandi.net), 4 (2007-8).

Meillet, Antoine, Les Origines indo-européennes des mètres grecs. Paris, 1923.

Nagy, Gregory, Comparative Studies in Greek and Indic Meter. Cambridge, Mass.: Harvard University Press, 1974.

Parry, Milman, The Making of Homeric Verse. ed. Adam Parry. Oxford: Oxford University Press, 1971.

Shive, David M., Naming Achilles. New York: Oxford University Press, 1987.

West. M. L., ‘Greek Poetry 2000-700 B.C.’, Classical Quarterly, 23 (1973), 179-92.

Whitman, Cedric H., Homer and the Heroic Tradition. New York: W. W. Norton, 1958.


by A. P. David.

©2009. All rights reserved. Use only with permission.