Off-beat Poetry: Rhythmic Games in Horace, Homer and Vergil

Nicholas Stone

Vergil (70 -19 BC) wrote his Aeneid at an average rate of about two and a half lines a day for eleven years (29–19 BC). Save for some unfinished lines,^[1]58 out of 9,896 hexameters are authorially incomplete, ending at different points within the line. we can reasonably suppose that every syllable in this epic poem has been put there because he wanted it there. Our job, as well as simply enjoying the fruits of his labour, is to work out why.

We can say this also of our engagement with other ancient writers. Vergil was rather extreme in the amount of time he expended on his composition, but literary work was a matter of great care for Classical authors more generally. Their works are full of literary devices. This is not to say that each device reflected a conscious intention to achieve a particular effect. A writer can write something and be pleased by the way it sounds or looks without having consciously thought through why it is pleasing.

The best we can do is to look and listen, and try to understand what effect (if any) is achieved by each apparent device – each deviation from an entirely normal and pedestrian use of language, or, conversely, each instance of apparent normality – and to make an educated guess at the likelihood that it was done consciously.

It usually is impossible to do this usefully in general terms; after all, what do we learn by reflecting that a simile always involves a comparison or that alliteration always attracts attention? To make progress, we have to consider particular cases. One of the things we have to grapple with as readers is that the same device can be used to achieve a different effect in different contexts. What remains consistent between these different contexts is simply that the device is in some way emphatic, by a means particular to that device.

Variation of the lengths of words in the cadence of the dactylic hexameter offers a useful example of the diverse uses to which a single device can be put.^[2]For more detailed information about how Greco-Roman quantitative metre worked, you could read this introductory article and/or have a more targeted look at the lectures and handouts of this course on the many forms of Greek and Latin metre. The dactylic hexameter, the line used for Greek and Latin epic poetry, has six feet, each either a dactyl (one long syllable and two short ones: – ᴗ ᴗ) or a spondee (two long syllables: – –). Most of the time, the last two feet (which we call the cadence)^[3]This word is derived from the Latin verb cadere, which means “to fall”. are a dactyl followed by a spondee: – ᴗ ᴗ – –.^[4]Sometimes the fifth foot may be a spondee. This happens more often in Greek than in Latin verse but is still uncommon (around 5% of Homer’s lines in the Iliad and Odyssey have a spondaic fifth foot). So the first line of the Aeneid ends primus ab oris: – ᴗ ᴗ – –. English phrases that can be used to remember this rhythm include “shave and a haircut”, or “strawberry jampot”.

In Classical Latin hexameters, it is much more normal for a line to end in a word of two or three syllables (what we call disyllabic or trisyllabic words) than in words of one, four or five syllables – or any greater number. This is because Latin words had an accent, which was based on stress (i.e. which syllable you stress more when saying the word): words of two syllables were stressed on their first syllable; words of three syllables were stressed on the second syllable if that syllable was long, or on the first syllable if the second syllable was short instead.

Given this natural pattern end of stress, ending a line with either a disyllabic or trisyllabic word would make it easier for the stress of the last words in the line to coincide with the “ictus” (that is, the rhythmical beat) which opens each metrical foot of the hexameter. In primus ab oris, the beats of the last two feet fall on the same syllables that carry the stress of the accent (on primus and oris).^[5]This fact is guaranteed for all disyllabic and trisyllabic final words because of the line’s preordained rhythm: all such words will have their accent on the penultimate (necessarily long) syllable, and any word completing the fifth foot will have its accent on the long syllable that opens that foot. Because the stresses of the metre and the accent occur on the same syllables, there are fewer potential syllables to emphasise (i.e. two rather than three or four); the phrase thus flows more smoothly and provides a sonorous and satisfying close to the line. Multiple stresses within a foot would seem conflicting, and the sound much choppier.

Because of this, line-endings that allowed the accent and ictus to coincide in the last two (or three) feet came to be seen as more elegant, and became more normal: for instance, they are much more common in Vergil, writing the Aeneid in the 20s BC, than in Lucretius, writing a generation earlier in the 50s BC. Over time, it became the case that ending a line with a word that was not disyllabic or trisyllabic would sound a bit odd.

What did poets do with this possibility of “sounding a bit odd”? It depends.

One of the most famous examples of a line ending in a word of one syllable comes from Horace, in his Ars Poetica (composed c.19 BC). Horace borrows an image from Aesop’s fables, of a mountain giving birth to a mouse (perhaps the oldest popular metaphor for an anticlimax?), and uses it to describe a poet writing an epic with a beginning that he then cannot live up to:

nec sic incipies, ut scriptor cyclicus olim:
“fortunam Priami cantabo et nobile bellum.”
quid dignum tanto feret hic promissor hiatu?
parturient montes, nascetur ridiculus mus.

(Ars Poetica, 136–9)

Don’t begin like this, as one writer of the epic cycles once did: “I will sing the fortune of Priam and noble war.” What could this promiser bring that would be worthy of so great an opening? The mountains will labour, and there shall be born a ridiculous mouse.

Horace underlines the extent of the poet’s failure in a few ways. As well as simply meaning someone who has given a promise, promissor is a Roman legal term for somebody who has undertaken a contractual obligation. We are nudged to associate the poet here with somebody who has taken on a debt which he now cannot repay. The word hiatus, as well as meaning a beginning, denotes an open mouth. We are encouraged to think of the poet standing there, jaws agape, like the mountains dividing, as the inadequate poetry flows forth.

But the climax of these images is the monosyllabic ending of line 139. The inadequacy of the poet’s offering is stressed by the appearance of the absurd image, the ridiculous mouse emerging from the mountains, whose appearance is delayed until the end of this line, and of this sequence of lines, as a punchline. We have to wait until the end for the full force of the criticism. But it is more powerfully emphasised than by its position alone.

ridiculus mus ends the hexameter: its rhythm is, as we would expect, – ᴗ ᴗ – –. But the monosyllabic ending means that there is no coincidence of accent and ictus in this cadence. Instead, there is a metrical stress on the usually unstressed last syllable of ridiculus, and a forced pause created by the break between words, before an unusual, accentual stress on the last syllable of the line, the one-word surprise, mus. This fragmentation of the rhythm at the line’s end draws our attention more to this word and its smallness – like the smallness of the literal and metaphorical things it denotes.

We can look to Vergil for a more abusive effect achieved by this device, of varying the lengths of words in the hexameter cadence. In Book 4 of the Aeneid, we meet Iarbas, a son of Jupiter and the king of Getulia in North Africa. He was a suitor of the Carthaginian queen Dido, before her tragic entanglement with Aeneas. Here we see him complaining to Jupiter about their relationship:

                                                …conubia nostra
reppulit ac dominum Aenean in regna recepit.
et nunc ille Paris cum semiviro comitatu,
Maeonia mentum mitra crinemque madentem
subnexus, rapto potitur…

(Aeneid 4.213–17)

…she has rebuffed our marriage suits and received Aeneas into her kingdom as master. And now that Paris with his half-man crew, with a Lydian band on his head and done up with dripping hair, obtains his booty…

By associating Aeneas with the cowardly, woman-stealing Trojan prince Paris and alluding to stereotypically eastern refinements of dress, Iarbas salves his ego by casting his love rival as a deviant and inferior figure. Aeneas is imagined dolled up like a perfumed seducer, rather than as a warrior, and is equated with the Trojan prince who devoted his attentions to chasing women, lounging behind the walls during battle, and fighting with a bow – a suspect, unmanly weapon since it could be used from a distance. If Aeneas is Paris, Iarbas in turn must be Menelaus – the wronged husband whose wife has been stolen away by a foreign playboy whom he could easily beat in a fair fight.

The greatest emphasis to this picture of Aeneas, the stereotypically unmanly eastern intruder, is given by Iarbas’ use of the words semiviro comitatu, “with his half-man crew”. Partly this is because of the literal meaning of the words. But partly it is because of the metre. For the preceding hundred lines, Vergil has given us only disyllabic or trisyllabic cadences. At 215, with semiviro comitatu, we have a tetrasyllable following another tetrasyllable.

The accentual stress of semiviro would already be a bit ungainly. A four-syllable word with a short penultimate syllable would normally be stressed on the antepenultimate syllable (so, semiviro); but because this is a compound of two disyllabic word elements, there may be a temptation to stress it as though it were two disyllabic words (so, semiviro). However you say it, it sounds wrong. And where Vergil has placed the word, with its last syllable beginning the fifth foot of the line, the main metrical stress of semiviro falls instead on its final syllable (semiviro). So, however wrong it sounds normally, Vergil has made it sound even wronger, and forced you to give a metrical stress to a syllable that could have no accentual stress at all.

This ungainly beginning to the cadence is followed by the oddity of another four-syllable word. In comitatu, the stress accent and ictus do coincide, on the third syllable (comitatu). But the fact that they coincide on the third syllable of a four-syllable word nevertheless stands out as odd, after a hundred lines that ended in words of two or three syllables – an effect that is doubly uncomfortable coming after the ungainly semiviro. The oddity of this pattern of stresses at the line end, falling upon words that describe men who are “half-men”, cannot help but make the line-ending seem to signify that same oddity: just as Aeneas and his crew are not proper men, this is not a proper ending to an epic hexameter. Something is wrong with the Trojans, and that is why something seems wrong with this line; something seems wrong with this line, so it reinforces Iarbas’ statement that something is wrong with the Trojans. A tetrasyllable at the end of a dactylic hexameter does not have to be a device; but here, it is.^[6]It seems that some Romans thought that a line ending with too many syllables would always sound effeminate – see, e.g., Quintillian, Institutio Oratoria 9.4.64–6. But there is room for debate over whether Quintilian’s generalisation would apply to all tetrasyllables or only to spondaic ones, given that all the examples he gives are spondaic. In any case, this sort of ancient judgment seems something that ought to be considered when working out the likely effect of a line rather than a proof of its effect one way or the other.

Finally, we may turn to Homer for another monosyllable of interest. In Greek, words had a tonal accent rather than a stress accent (i.e. the accent indicated a rise in pitch rather than an increase in the force with which a syllable would be pronounced). So, without any aesthetic pressure to accommodate a stress accent to the metrical ictus, the tendency to have line endings of disyllabic or trisyllabic words was much less acute. In Homer, it is quite normal for the line-ending to be occupied by a five-syllable epithet or name, e.g. Penelope’s: Πηνελόπεια (Pēnelopeia).^[7]As found at, e.g., Od.21.311: τὸν δ᾽ αὖτε προσέειπε περίφρων Πηνελόπεια. Leaving metre for a moment to attend to alliteration: one might mark the six sounds of “p” in this verse (underlined, five πs and one φ; only the last two, in Penelope’s name, are on the ictus). Four of them are either in Penelope’s name, or in her epithet, περίφρων (“prudent”), which might well have become associated with her in the first place because it alliterates. Why mention this line? Partly as a cautionary tale to contrast with the more probable examples of literary effects above: it is often hard to say whether there is anything intentional going on. Aside from the (fossilised? subconscious?) attraction of sound and sense in the epithet-name alliteration, it is likely enough that the “p” sounds in this line are not seeking to do anything. And yet it is just about possible that they are. Such small additional attention as is placed on Penelope’s epithet by the two prior “p” noises may draw a bit more attention to the contextual irony of this epithet in the light of what she is about to say – that the beggar in the hall (who, unbeknownst to her, is her husband Odysseus in disguise, shortly to reclaim her) would not possibly think of vying for her hand. Nevertheless, things did occur that were less common. One of these was the use of a one-syllable word at the end of a line. It did not necessarily have to mean anything in particular, but it might occasionally be suggestive.

ὡς δ᾽ ὑπὸ λαίλαπι πᾶσα κελαινὴ βέβριθε χθὼν
ἤματ᾽ ὀπωρινῷ, ὅτε λαβρότατον χέει ὕδωρ
Ζεύς…

(Iliad, 16.384–6)

hōs d’ hupo lailapi pāsa kelainē bebrīthe khthōn
ēmat’ opōrīnōi, hote labrotaton khe-ei hūdōr
Zeus…

As when all the dark earth is heavy beneath a storm on an autumn day, when Zeus pours the most furious rain…

Here, we have the beginning of a long simile at the point when Hector’s chariot speeds away from Patroclus, who is battling the Trojans while wearing the armour of Achilles. The direct point of comparison in the simile is between the oppressed, foreboding heaviness of the earth under an impending flood from Zeus that will wash away the unjust, and the cry of Hector’s mares as they bear him away.^[8]Iliad 16.384–93. In these initial lines, the image of the wrathful Zeus inflicting rain increases the terror of the present slaughter of the Trojans by Patroclus; the same comparison and the sense of foreboding created by the image of a storm welling up acts also as a reminder of both Hector’s and Patroclus’ eventual meeting, and their respective fates under the eyes of Zeus.

There is a lot going on here. But let us focus only on the words βέβριθε χθὼν: the earth is heavy. This is a spondaic line ending (– – – –), which, as we have noted, is not so rare in Homer. But the use of only one word at the end of the line is rather more unusual. It draws the focus somewhat more to itself. In this context, where it is the subject of the verb “is heavy”, a monosyllabic conclusion as the last of a sequence of four long syllables seems to mimic the heaviness in the earth described. It is like a thud. In such a metrical and thematic context, the succession of harsh thetas and chis (βέβριθε χθὼν) is perhaps suggestive of collision, as something falls – rain, a hoofbeat, a weapon, a body, fate…

It is harder to say any of this with confidence than it was regarding the effects of Horace’s ridiculus mus or Vergil’s semiviro comitatu – not least because it is harder to say with confidence that there is a self-conscious process of literary composition at work in each Homeric line compared with each Vergilian one. But perhaps it should increase our confidence that there might be some significance in the way the ending of this line has been shaped, that the previous substantive monosyllabic line-ending in the poem comes some 80 lines earlier, where another element, πῦρ (pūr, fire), is at work (in that case a real fire, corresponding to Zeus’ lightning in a preceding simile). But, of course, even if it is the case here that a monosyllable at the end of the line has enacted a heaviness described in the line, it does not mean that we should suspect heaviness whenever we see a line ending with a monosyllable.

This toying with monosyllables and tetrasyllables is just one example of the diversity of effects that can be achieved by attention to one of the species of devices available to Classical authors. Each device can have a variety of effects. The degree of certainty we can feel in each case as to whether the effect is really there (rather than a trick of the light) and, separately, whether it is intentional, must vary with the evidence. For my part, I will consider only two and a half lines a day for the next eleven years, just to be on the safe side.

---

Nicholas Stone studies Law at Harris Manchester College, Oxford. He has written various Latin poems for Antigone, which can be found here, here, here, and here.

---

Further Reading

Metre has an unjustified reputation for being complicated. Some aspects of it are quite complex (for instance the analysis of Greek choral lyric, where the form of the poem can vary hugely, both within and between poems), but some are quite straightforward, like learning how to read a dactylic hexameter with the right rhythm. This is just a matter of patience and practice, and it makes reading the poems much more fun.

David Butterfield’s introductory essay and lectures on metre, hosted by Antigone, are a wonderful way to learn more about this subject. Armand D’Angour’s “Mnemonics for Metre” are ideal for memorising the different rhythms quickly, before trying them out on the Classical texts once you’ve got the hang of them.

The introductory books by D.S. Raven, Greek Metre (Faber & Faber, London, 1962) and Latin Metre (Faber & Faber, London, 1965) are very useful for getting the basics of the development of different metres. On Greek metre, M.L. West’s Greek Metre (Oxford UP, 1982) remains the standard scholarly work. For a more in-depth look at Latin metre, see L.P. Wilkinson’s article “The Augustan Rules for Dactylic Verse”, Classical Quarterly 34 (1940), 30–43, Maurice Platnauer’s Latin Elegiac Verse (Cambridge UP, 1951), and Llewelyn Morgan’s Musa Pedestris (Oxford UP, 2011).

On the Latin dactylic hexameter in particular (and particularly as used by Vergil), see also the splendidly detailed work of S.E. Winbolt, Latin Hexameter Verse, An Aid to Composition (London: 1903). It is worth consulting (and trying to argue with) this book, as well as 19^th– and early-20^th-century manuals of verse composition in general, even if you are not interested in writing your own poems in Latin or Greek (although doing so is probably the most fun way to try, however imperfectly, to get a practical perspective on the poetic forms).

Jasper Griffin’s Homer on Life and Death (Oxford UP, 1980) and Richard Rutherford’s Homer: Greece and Rome New Surveys in the Classics (Cambridge UP, 2013, 2^nd edition) both contain useful explanations of Homer’s style and of the role of literary devices in structuring the Homeric poems; both point the way to much further reading.

Much ancient poetry (particularly Greek lyric and epic) was originally sung rather than read. For detailed explorations of how this worked and what it meant for the poems as poems, see Armand D’Angour and Tom Phillips’ Music, Text and Culture in Ancient Greece (Cambridge UP, 2018), as well as some articles on Antigone.