Let’s take stock of what’s been said. I was stopped short in this performative journey through the Odyssey by the problem of how to think and talk about Persephone. Perhaps there is a significance hidden to me in such a blockage. Homer himself gives us next to nothing about the goddess, apart from her epithets such as ἐπαινή ‘dreaded’ or ἀγαυός (‘noble’? ‘handsome’?). While I feed happily off internal reverberations, it does seem, as in the case of Persephone, that there are definite allusions to figures and stories beyond and behind the stage of Homer’s foreground. This has led me to consider, in the first case, that allusion is itself, in species, a kind of musical resonance, like memory; and then to focus on the figure performing solo on Homer’s stage like Odysseus, without a musical instrument, wielding a staff, and the ways he conjures presences in the course of his storytelling. The most obvious of these is when he takes on the persona of a protagonist, and speaks hexameters in his or her voice. At least as visible, however, must be the use of his body and especially his staff. There may be evidence that the Homeric poems were composed for such a thespian performer, a rhapsode or rhabdode—the latter variant means ‘staff-’ or ‘wand-singer’—rather than a bard playing a sort of lyre, a tableau described several times in Homer.

The depiction of these bards strongly suggests that Homer’s own artwork participates in a musical tradition. Today let us focus on Homer’s harmony. Insofar as musicology studies such historical transmissions, including in the relation of written notation to physical and auditory performance, there should be insights there, in musicology, to be mined in the investigation of the transmission of Homer. Something, at any rate, must free us from the cultivated opacity of the preferred academic debate which surrounds orality and literacy, with its lingering whiff of the noble savage. I suppose the distinction between orality and literacy is one that makes us feel intellectual, without having actually to think. There is in fact reason to believe that alphabetic writing arose in Greek as a musical notation, literally a way to record Homeric μουσική, for which a syllabary such as the Mycenaean was inadequate.

Please also note, however, that there is no reason whatsoever, and never has been, to believe that Homer was an oral poet serving in some sort of tradition of oral poetry, no matter that all academic authorities you may see, hear, or read, will usually now assume these things as given, and to be historical facts. These authorities will even soil their intellects by pronouncing certain dates for ‘Homer’ measured in Centuries B.C.E., though they have no background in scientific chronology, and are wont to invent a whole swath of an ancient Greek timeline, ominously called a ‘Dark Age’, by denying the stratigraphic facts of empirical archaeology. Instead they assign dates to artefacts in the Greek Mediterranean (outside of Cyprus! by reason of some code of the guild) by coordinating with the notoriously unscientific, jury-rigged, humanly motivated construct called Egyptian chronology. This coordination is the only source of a Greek Dark Age of hundreds of years, separating the Mycenaean ware and writing in Linear B, dated via Egypt, from the immediately contiguous and even overlapping archaic strata in Hellenic-speaking lands.

But surely the most damning deficit in the present oral orthodoxy about Homer has to do with its direct neglect of Homer’s text as we have received it from the Alexandrian Greeks. This text is visibly and manifestly a musical score, as surely as Homer’s artwork was universally described in the ancient world as music (μουσική) and, within Homer himself, as song (ἀϝοιδή). The entire stream of Homer’s composition is, in the descendent manuscripts, divided visually into what we now call lexical words and word-like groupings, and each of these ‘words’ is marked with an accent which notates a pitch contour of both a rise and a fall; different ancient grammarians have indicated that the interval of this contour may have been either a fourth or a fifth. Usually the whole ‘contonation’ takes two or at most three syllables to complete, limited to the last three syllables of the word, although the accent sign only marks the syllable within which the rise in pitch begins. Note that the fall goes unmarked except in the case of those vowels marked with a circumflex, in which case the whole uprise and downfall, a ‘contonation’, happens over the length of that vowel. But the comparison with Vedic strongly suggests that the down-glide also occurred in Greek, on the unmarked syllable or syllables immediately following the marked rise. I have shown how either the rise or the fall in pitch becomes dynamically prominent, or ‘stressed’, depending on the duration of the syllable on which the falling pitch lands. Hence amidst each word-level pitch contour there arises a most prominent syllable which can dynamically accent a beat in the underlying dactylic feet, LONG-short-short, so that syncopation can break up the reinforcement of that beat and create musical rhythm, rather than the robotic metronome of metrical scansion. Punching the metrical downbeats, what classicists call ‘scanning’, gives you this:

ἄνδρά μοι ἔννεπε, Μοῦσα, πολύτροπον, ὃς μάλα πολλά

while intoning the pitches and stresses gives you

ἄνδρά μοι ἔννεπε, Μοῦσα, πολύτροπον, ὃς μάλα πολλά

In a comparable way, English stresses occasionally syncopate the iambic pentameter to turn monotony into rhythm. Nor sháll deáth brág thou wander’st in his shade.

There emerges out of my work, for Homeric and classical Greek as well as Latin, a Law of Tonal Prominence. Let’s illustrate it by example. These are the opening three words of the Iliad:

μῆνιν ἄ[ϝ]ειδε θεά,

First off, the digamma is there in brackets because it does not appear in our inherited texts, but the presence of its -w- sound in Homer’s composition has been inferred from a number of angles, such as comparison with Indo-European cognates, fitting the Alexandrian syllables to the dactylic hexameter, and saving Homer from hiatus, the clashing of vowel upon vowel. This latter avoidance was anciently ascribed to Homer, but in our present text his second word is guilty:

ἄ-ειδε

Fortunately we can reconstruct a form:

ἄϝειδε (pron. a-wei-de)

I have suggested that avoiding hiatus is not the expression of some esoteric aesthetic or grammarian’s fetish, but a feature sprung from the concrete reality of singing:

The ‘avoidance of hiatus’ is in fact the way grammarian philologists register a phenomenon that was actually caused by a cantor’s need, and desire, to articulate and voice the vowel of any sung syllable with a distinct, specific, initial consonant. It is not about separating vowels like naughty sheep, but articulating them as the bearers of melody in song.

The story of the digamma illustrates that language is something we inherit, but that even among the attentive and the reverent, something gets lost in the transmission; and at the best of times, even without the damage of catastrophe or upheaval, we nevertheless fail to pass on our language unchanged. One witnesses more changes the longer one lives, in real time, and increasingly they become topics of whining conversation among peers in age. Sounds change, meanings change, accents shift. British and American stress habits coexist though we speak the same language. Where Americans disappear the first ‘o’ in LAB-ra-TO-ry and stress the second, the British do the opposite to their ‘o’s when they stress the antepenult, lab-OR-a-try, while both conserve the spelling ‘laboratory’.

Tracing the history of such changes is a forensic work with uncertain results. The consensus is that the Proto-Indo-European accent was similar to the Vedic one, a tonal contonation like the one we find in Homer, rising and then falling in pitch, within a syllable or on adjacent syllables. But there is evidence in some Greek forms of an historical stress accent in Greek, earlier than Homer, which weakened unstressed vowels. All the same, the pitch accent in Homeric and later Greek appears to have the same nature as the Vedic tonal accent, which does not result in the weakening of adjacent vowels. And unlike the descendants in Greek, Latin, and Sanskrit, the Vedic accent was not recessive, that is, tied to the last syllables of a verbal unit. Instead its high pitch point appeared to be free to occur anywhere in a polysyllable.

The upshot is that when the ancestor of a Greek word was accented too early, the onset of the accent would have had to be shifted forward to become Hellenised. That is, the Greek accentual habit, whatever its source in the mysterious workings of the will or the brain or culture, imposed its ways on inherited words. But its influence was not tyrannical; if the inherited onset of the contonation already occurred within the final three syllables of a word, it was allowed to maintain its place rather than be forced to recede. Hence we have attested μητήρ, μητέρος, ‘mother’ or ‘mother’s’, rather than recessive μήτερος. Such inherited accent places are said to be ‘lexical’ in that they belong to the lexeme itself and are not forced, as when singing the finite verbs, by the Greek recessive proclivity. The circumflex in μῆνιν is lexical in this sense, although it happens to conform to the recessive rule. In most of the Greek dialects only the verbal forms, with few exceptions, show a recessive onset of the accent (the part with rising ...

Homer’s Harmonia: In the Beginning ... was the Musical Word