6. The Heavy Groan and the Heavy Accent1
Gregory Nagy discusses a phrase from the Venetus A manuscript that includes a third-foot cadence, τὴν δὲ βαρὺ στενάχων (or στεναχῶν), the first portion of Iliad I.364: ‘Groaning heavily [he addressed] her’. His discussion is one that serves to illustrate a) by omission, the import of applying the new theory of the accent in interpreting manuscripts, and in particular the principles underlying what should be called the proposition of Venetus A; and b) the precious value of the resources still contained in that manuscript, for those capable of Nagy’s excellence at forensic work. This forensic work, revealing the true source of the reading with the circumflex (στεναχῶν), underscores the propositional nature of the Venetus A text, where lemmas are given in support of the final rendition for select passages of the main text, and arguments are made in the scholia, by way of instances and examples, for the lemmas on which the passages’ diction and accentuation are based.
The proposition made by the Venetus A is that the text of Homer for the first part of I.354 is τὴν δὲ βαρὺ στενάχων. The lemma on which this proposition is made is an authoritative text which appears to read τὴν δὲ βαρυστενάχων. All the focus is on the final accent. The primary scholium makes an argument for the lemma on synchronic grounds, implicitly arguing against the prosody βαρυστεναχῶν as a contraction of βαρυστεναχέων, because we would then elsewhere in Homer see the spellings ἐστεναχοῦντο in the verb and στεναχοῦντος in the declension of the participle—the οῦ in each penult reflecting the contraction of έ + ο—instead of the cited ἐπεστενάχοντο, together with στενάχοντος and στενάχοντι. Nagy attributes this argument to Herodian.
But the argument for the lemma gets a good deal more interesting in the secondary scholia, which also seek to ‘prove’ the lemma; Nagy makes a convincing case that Aristarchus is the influential figure behind Herodian and the other scholiasts to Venetus A in making the case for the chosen accentuation. But in the Bt scholia, and a scholium reported on Odyssey 5.83, a case is made for a form στεναχή which was, by inference, the reading of Aristophanes of Byzantium via a student named Dionysius of Sidon. (Aristarchus was also such a student, more senior; he became, it would seem, the successor to Aristophanes in authority at Alexandria.) Consistent with a form στεναχή comes the alternate reading of I.364, preferred by the earliest authority on accentuation: βαρυστεναχῶν. Nagy shows that Aristophanes read a form στεναχῇσιν in the Odyssey. There is therefore a diachronic case for reading στεναχῶν, or βαρυστεναχῶν, at line 364, which in my view—and, one suspects, Nagy’s as well—trumps the proposition made by Venetus A for this line, together with its supporting synchronic lemma.
The marginal scholia, however, provide additional support for the lemma, and while they do not in my view win the day, they do provide a most valuable insight into Aristarchus’ biases as we look through the Venetus A, in some cases, to the Homer of Aristophanes of Byzantium—the parent of its graphic accentuation. In the margin we read, ὅτι τὸ βαρυστενάχων κατὰ βαρείαν τάσιν, ‘because “βαρυστενάχων” is [pronounced] according to the heavy accent.’ Nagy comments: ‘When the text speaks here of barytone accentuation, it has to do with the avoidance of placing an accent over the final syllable of a word.’2 Nagy is not alone in interpreting a phrase that literally means ‘heavy accent’ (βαρεῖα τάσις) as meaning ‘unaccented’ in Greek grammarians’ usage. This absurd violence to the usage of Plato, Aristotle and the grammarians, about the word ‘heavy’ (βαρύς), is endemic in the prevailing teaching about accent in schools.
The new theory at last makes sense of the word. It is a positive term—‘heavy’—in no sense a privative. Its meaning in prosody is akin to the one it has in describing Achilles’ groaning in the sympathetic resonance of this very line. The heavy accent is characterised by a down-glide in the voice over a long or heavy syllable; it is signalled by the preceding acute. Clearly this positive description by the scholiast does not refer to the effect of the rise in pitch on the alpha in this word; it describes the phenomenon of the heavy fall in pitch on the long omega that follows. This is the phenomenon that reinforces the ictus on the thesis of the third foot. What the scholiast calls ‘the heavy accent’ is in my new parlance a ‘post-acute barytone’ on the unmarked final vowel of στενάχων.3
The new theory, on analogy with the use of svarita in the description of Vedic, also treats the circumflex as a barytone, and the reciting of a circumflex could be just as evocative of a moan or groan in this context. The scholiasts distinguish both the sign and the break in the sound, it would seem, in referring to Aristophanes’ reading as περισπωμένη, ‘bent-over’ or circumflex, but in the view of the new theory, both readings reinforce the ictus on the same syllable. This is not therefore a disagreement about which syllable is most prominent or accented, as it might appear to be visually (between στενάχων and στεναχῶν). Both prosodies reinforce the same element that they need to reinforce, the longum of the third foot in the line. This is an insight of the new theory. The disagreement is, explicitly, over the quality of that reinforcement: circumflex (περισπωμένη) versus barytone (κατὰ βαρεῖαν τάσιν).
Nagy draws on a fascinating contribution from the bT scholia to this passage, which seems to offer a reason for Aristarchus’ choices; it would seem that generally it is his choices behind the lemmas that justify the Venetus A text.4 Note in particular this comment from bT, justifying the choice of post-acute barytone over circumflex in this instance:
The Sidonian circumflects; for by circumflecting he also cites “ἁδινὰ στοναχῆσαι”.
(The ‘Sidonian’ is Dionysius of Sidon; Nagy argues, with evidence, citing a scholium to Odyssey 5.83, that the circumflex he prefers for this word is due to Aristophanes of Byzantium, who knows the form στεναχή.)
But Aristarchus sings it heavy [βαρύνει]; for most of the transitions [κινήματα] arise based on the barytone [ἀπὸ βαρυτόνου]:
He then supplies two examples of such ‘movements’, one each from the Odyssey and the Iliad:
βαρὺ δὲ στενάχοντος ἄκουσεν (8.95), ἐπεστενάχοντο δ’ ἑταῖροι (IV.154).
Nagy translates the scholiast’s last clause ‘since inflections [κινήματα] happen for the most part by way of starting from barytone accentuation.’ As we mentioned, ’barytone’ under the modern convention effectively means ‘without an accent mark’ on the ultima. Inflections of the verb generally show recessive accent, and so ‘for the most part’ verb forms do not show an accent mark on their final syllables. But this is obviously not the case with contract verbs. So what exactly is the case being made here? Is the possibility of the form στεναχῶν, from στεναχέω, being rejected simply because contract verbs constitute a somehow undesirable minority among verbs? Whether or not the scribe who has transmitted this scholium understood what was being said, it may be that Aristarchus intended an argument in support of the lemma that was less naive in opposing his own teacher.
It seems to me possible that the sense of ‘inflection’ Nagy gives to κινήματα, based solely on the usage of the Roman-era grammarian Herodian, may mask a more musically interesting bias on the part of Aristarchus. If it is most ‘movements’ that depend on the barytone, there may be an observation here about the characteristic prosodic movement of the hexameter. Here is my translation and discussion of a passage from Plato’s Timaeus about the κίνησις, the movement of accent and rhythm from disagreement to agreement (80a ff.), which may inform our reading of κινήματα in the scholiast:
καὶ ὅσοι ϕϑόγγοι ταχεῖς τε καὶ βραδεῖς ϕαίνονται, τοτὲ μὲν ἀνάρμοστοι ϕερόμενοι δι’ ἀνομοιότητα τῆς ἐν ἡμῖν ὑπ’ αὐτῶν κινήσεως, τοτὲ δὲ ξύμϕωνοι δι’ ὁμοιότητα. τὰς γὰρ τῶν προτέρων καὶ ϑαττόνων οἱ βραδύτεροι κινήσεις, ἀποπαυομένας ἤδη τε εἰς ὅμοιον ἐληλυϑυίας αἷς ὕστερον αὐτοὶ προσϕερόμενοι κινοῦσιν ἐκείνας, καταλαμβάνουσι, καταλαμβάνοντες δὲ οὐκ ἄλλην ἐπεμβάλλοντες ἀνετάραξαν κίνησιν, ἀλλ’ ἀρχὴν βραδυτέρας ϕορᾶς κατὰ τὴν τῆς ϑάττονος ἀποληγούσης δὲ ὁμοιότητα προσάψαντες μίαν ἐξ ὀξείας καὶ βαρείας ξυνεκεράσαντο πάϑην, ὅϑεν ἡδονὴν μὲν τοῖς ἄϕροσιν, εὐϕροσύνην δὲ τοῖς ἔμϕροσι διὰ τὴν τῆς ϑείας ἁρμονίας μίμησιν ἐν ϑνηταῖς γενομένην ϕοραῖς παρέσχον.
[We must pursue] also those sounds which appear quick and slow, sharp [ὀξεῖς] and heavy [βαρεῖς], at one time borne in discord because of the disagreement of the motion [κίνησις] caused by them in us, but at another in concord because of agreement. For the slower sounds overtake the movements of those earlier and quicker ones, when these are already ceasing and have come into agreement with those motions with which afterwards, when they are brought to bear, the slow sounds themselves move them; and in overtaking they did not cause a disturbance, imposing another motion, but once they had attached the beginning of a slower passage, in accord with the agreement of the quicker one, which was fading, they mixed together a single experience out of sharp and heavy sound, whence they furnished pleasure to the mindless, but peace of mind to the thoughtful, because of the imitation of the divine harmony arisen in mortal orbits.
There is something for everyone here, the mindless and the thoughtful, the pop and the classical, in the poetic ‘motion of movements’ of the Muses. The dance circling with retrogressions imitates in sympathy the apparent motion of the outer planet-gods. From my analysis:
… it is natural to read ‘similarity’ here … as a correspondence of quick to sharp and of slow to heavy. Such a correspondence constitutes ‘agreement’. ‘Disagreement’ would arise out of the opposite collocations … When a term subsumes a pair of definitive contraries—as, for example, ‘number’ in relation to the ‘even and odd’—Plato sometimes treats the pair as synonymous (or metonymous) with the term itself (see, e.g., Laws 818c, Epinomis 990c). Hence the pairs here seem likely to refer to what he elsewhere calls ‘rhythm’ [quick/slow] and ‘harmony’ [sharp/heavy]. In particular, in the context of the motile internal dynamism of a rhythmic foot, quick would most naturally refer to the arsis, which contains one or two shorts, while slow refers to the long thesis. As one considers the dominating influence of the slow and heavy sounds in the process described, a special weight may be given to the disagreement or ‘dissimilarity’ arising out of the conjunction of heavy and quick, where a heavy sound occurs in the arsis, as also the sense of agreement or ‘similarity’ produced by a heavy sound where it is supposed to be, in the thesis. Such disagreement and agreement is understood as belonging not to the sounds themselves, but to motions produced by the sounds ‘within us’. Later in the passage sounds are said to ‘move motions’ (κινοῦσιν κινήσεις). It is not clear whether these motions are understood to be entirely internal, or whether a literal reference is being made to orchestic performance. In a heightened state of poetic transport, perhaps the distinction becomes moot.
I then call attention to the granular interaction between accent and rhythm in the description, which appears to reflect the distinctive inflection points of an hexameter line:
There appear to be two points of dynamic moment in this description: the first when the slower sounds ‘overtake the movements of those earlier and quicker ones’; the second when they attach ‘the beginning of a slower passage’. At first it would appear that the interaction occurs entirely within the realm of rhythm and metre; slower sounds ‘overtake’ quicker ones. But in so doing, the narrator says, they have ‘mixed together a single experience [μίαν πάθην] out of sharp and heavy’. The interaction of harmony and rhythm begins at a trot in disagreement; then subtly turns, as at the cadence of the caesura, where slow sounds first ‘overtake’ the motion and come to a point of agreement. Then comes a new beginning, as at the diaeresis, leading to euphonic agreement in the coda … It is emphasized that the overtaking and the new beginning do not introduce a disturbance; rather, the new passage is ‘on the terms’ of the agreement reached in the earlier quicker passage. It would seem, therefore, that an agreement reached at the caesura becomes fully confirmed in the coda.5
This is without question a very difficult passage of Plato to interpret. All the same, it ought to be applied to the scholiast’s use of κινήματα in relation to his examples. I comment on the Timaeus passage:
[It] is … a[n] attempt by a native speaker without recourse to technical terms from a dead language, at describing the syncopation and the accentual cadence of verse, both the phenomena themselves and their physical effects … Our vertical sense of harmony draws a different kind of unity out of sound and time; but the horizontal, rhythmic sense of a melodic cadence, of disagreement seeking agreement in cycles of accent and rhythm, is still vital in western musical discourse.6
It does seem that the reason for choosing στενάχων over στεναχῶν attributed by Nagy to the scholiast, who in turn is supposed to be explaining Aristarchus—the fact that most verbal inflexions do not involve contract verbs—is extremely weak. If the idea is in fact that most ‘movements to cadence’ (κινήματα) within the hexameter line involve a post-acute barytone reinforcement, and rarely a perispomenon one, with the implication of a musical preference at the cadence for the barytone contonation over the monosyllabic final circumflex, Aristarchus could be seen to be exercising his taste in avoiding the rare implementation of a terminal circumflex in this role—in line with what seems to be Homer’s own taste in the matter. The scholiast points to two cases where forms of στενάχω reinforce the fifth thesis (στενάχοντος, ἐπεστενάχοντο), the beginning of the coda in lines without a diaeresis; in I.364, however, the line in question, βαρὺ στενάχων lands on the third thesis, which is the kind of musical reinforcement that is the true cause of the famous caesura. It is indeed rare for this position to be reinforced by a final circumflex like στεναχῶν. But modern taste will favour the diachronic argument, the unusual form, and the pioneering practice and seniority of Aristophanes of Byzantium over Aristarchus. Βαρυστεναχῶν.
It must be said, however, that it is not altogether clear why Nagy takes an interest in this disagreement between Aristophanes and Aristarchus, decided in favour of the latter in the proposition of Venetus A. Nagy makes no secret that he does not credit this manuscript’s data as a source for earlier versions of either Homeric prosody or melody. But it would be strange if he intended instead that Aristophanes was somehow a source superior to Aristarchus for knowledge of the prosodic or melodic patterns of a supposedly oral original. All the same, he appears to see some advantage to his argument in its opting in one passage against a reading of the pioneering Alexandrian himself, the Hellenistic Aristophanes of Byzantium. Perhaps the aim is to discredit all these late Alexandrian protagonists in the Homeric story, to clear the field for the oralist poetics of a precedent, imagined music and language of an oral tradition?
But Aristophanes is responsible, bless him, for the distinction between circumflex and acute on long vowels. This concretely descriptive distinction is enough to ground W. S. Allen’s comparison of the ancient Greek accent to the Vedic contonation, which extends, I have shown, also to the classical Latin accent. Neither the ancient grammarians, nor the Alexandrian Aristarchus, nor modern grammars of ‘Ancient Greek’ see fit to dispute this distinction. It is a distinction about which mora, of the two in a long vowel, that the contonation rises in pitch. The consequence is different sounding ways in which a circumflex (perispomenon) or barytone (post-acute) or long acute syllable each reinforces the strong position in a metrical foot. At the third thesis of the hexameter, it is almost always the first two options that apply: Homer’s taste, and Aristarchus’ judgement, appear to favour the post-acute barytone over the circumflex in this position.
Aristarchus would not be the first, however, to try to turn Homer’s predilections into rules that he must then be made to follow. Thank goodness the East Roman ethos was one that did not rewrite history in terms of contemporary practice, including our own contemporary, data-free, folk-patronising fixations on oral composition, but rather attempted to preserve what it had received, demonstrating with citations in support of lemmas, as succinctly, elegantly, and best it could. This allows that we can make our own aesthetic judgement between Aristophanes and Aristarchus, and about far more besides, than is dreamt of in their philosophy.
The ultimate paradox about Nagy’s purpose lies in the fact that if he wants there to be fixed melodies associated with the words, phrases, or lines that tend to be described in oral theory as ‘formulaic’, they are already there! As fixed as anyone could want in the East Roman writing system, anchored to the individual words in their specific order. This is the legacy of the received tonal-accentual marking system. It is precisely the repetition of phrases with a fixed pitch pattern which highlights that pattern outside of the phrase’s semantic registration, as a melody. In repeating it without changing it, one starts to sing the thing rather than merely say its name.
As I have written, one would as soon explain repetition in music as wetness in water. Repetition awakens the musical power of words, sometimes in favour of, or despite, their simple semantic intention. It is the height of perversity to see phrases that are repeated in Homer, in the way of oral theory, as pre-existing metrical ‘building blocks’. This view prevents their ever registering as musical events—even, in the practice of some scholars and translators, as semantic events. Filler is as filler does. It is only when they are first repeated that phrases become musical: I speak of ‘choral signifiers’ as a uniquely Homeric resource in storytelling.7 Homer’s choral signifiers can do more than can separated words, and no less.
This is not to say that the first time Homer uses a phrase he later repeats, that she is inventing the thing out of the whole cloth. It is not likely, for example, that Homer invented ἄναξ ἀνδρῶν as a title, usually for Agamemnon in the context of the Iliad, and almost always sung immediately after the cadential word break in the third foot of the hexameter. But in Iliad I.7—when it is first used in the poem, and only on this occasion—Homer splits the phrase before and after the cadence; the caesura falls after ἄναξ ‘lord’, throwing ἀνδρῶν ‘of men’ to the other side of the cadence. This split focusses light on Agamemnon’s status in the conflict, a power among men contrasted with the divinely ‘radiant’ Achilles. Thus we know we are in the hands of a composer alive to the possibilities: this is a poet actively engaging with what has been given in this traditional title, ‘lord of men’, rather than a worker slotting it into place in the assembly line.
Groan with me, about what the gods call summoning names, but men call metrical formulas: βαρυστεναχῶν!
Originally from A. P. David, ‘Singing Homer’s Spell: The Disyllabic Contonation and the Proposition Made by East Roman Manuscripts’, Dramaturgias: Revista do Laboratório de Dramaturgia da Universidade de Brasília, 19:7 (2022), 805-66; since revised.
Gregory Nagy, ‘Traces of an ancient system of reading Homeric verse in the Venetus A’, in Casey Dué (ed.), Recapturing a Homeric Legacy, Cambridge MA: Center for Hellenic Studies, 2009, 147.
The practice in Venetus A is, I believe, true to the original intent of Aristophanes of Byzantium, and this is what I follow in performance. To reiterate: the acute signifies the vowel mora of the rising pitch—on a long vowel always the second one; circumflex signifies rise and then fall on the same long vowel, with the fall dominating; grave means suppression of the rise due to the terminal boundary of a word or phrase. Acutes are prominent prosodically only when the following syllable is light; if the following syllable is long or closed, the combination of pitch change with duration renders it the most heavy: a post-acute barytone.
see Nagy 156.
David, The Dance of the Muses: Choral Theory and Ancient Greek Poetics, Oxford: Oxford University Press, 2006, 89-91; for a different reading see Andrew Barker, Greek Musical Writings, ii: Harmonic and Acoustic Theory, Cambridge: Cambridge University Press, 1989, 62.
Ibid., 91-2.
Ibid., 138-71.