The Long and the Short of Latin Poetry

David Butterfield

The written word can be divided into two realms: that of prose and that of verse. But where exactly does the boundary run between them? That’s simple, you may think. Prose is just words, whereas verse is words that have some sustained rhythm to them. True enough – but it’s a little more complex than that. Not all verse is poetry, and not all poetry is verse.

Take a random proverb (“While the cat’s away, the mice will play”) or advertising slogan (“Once you pop, you can’t stop”). Both have a rhythm and feel that distinguishes them from prose, but poems they surely are not. Likewise, many a modern poem has little to mark it out as verse, other than its artistic (and often ideosyncratic) line-division on the page: set rhythms or rhyme schemes are no longer a poetic prerequisite. Music lyrics have their own rhythmical schemes, and are evidently a kind of verse – but does this make “She’ll think I’m superman, not super minivan” or “penalty shootouts are highly overrated!” or “I turn an injury into tumiditypoetry? I’ll let that be your call.

If you like…

Where then do these words come from, this ‘prose’, ‘verse’ and ‘poetry’? The answer is Latin: prose (from prosus, the shortened form of pro-versus) is “straightforward” speech; verse (versus) is something “turned” into a particular form or shape; and poetry (poēsis) is the work of an artist (poēta). In fact, this last pair of words comes directly from the Ancient Greeks (ποιεῖν / poiein = “to make or do”), whose culture drove so much of Rome’s own artistic endeavours. And, despite the myriad complexities of these two cultures, the line drawn between prose and verse was obvious and impermeable.

So let’s open the door a little into Greek and Roman ‘metre’ – the system that measured out their verse. To many newcomers this is a world of bewildering jargon, of terms that seem better suited to the laboratory – ‘synizesis’, ‘Reizianum’, ‘telesillean’ – or the zoo – ‘lesser asclepiads’, ‘limping iambics’ and the occasional ‘catalexis’. But the basic principle of the system is astoundingly simple. Each syllable of a word is classed as ‘long’ or ‘short’ (or, more properly, ‘heavy’ and ‘light’), and each metre is a set pattern of these long and shorts.

A brace of images from Carlo Carlucci’s groundbreaking survey, A Field Guide to Greek Metre (Atypical, Berkeley, CA, 1978).

To begin with English, let’s consider the shortest and longest words of one syllable: “a” is scarcely a split-second blurt when it appears in a phrase such as “in a bit”, but “strengths”, the longest monosyllable in the language,[1] takes an appreciable time to sound out. We thus have a clear pair of short and long syllables, but it’s anybody’s guess how to class middling words, such as “crack” or “blog”: if they have to be either longs or shorts, who makes the call and how do they know?

Thankfully, this is not how we do things in English poetry. But the Greeks and the Romans, when faced with similar problems, decided to operate with a false binary, whereby any syllable of any word was marked either short or long. Simply put, their rules were as follows:

  • Every syllable in Latin and Greek has a vowel at its heart. If that vowel is itself pronounced long, or two vowels are combined in one sound (a ‘diphthong’), that syllable is long (typically marked with the longum ). So, if we transferred this into English, bone, thigh, feet and heart would all be long.
  • If the vowel in the syllable is instead short, that syllable is short (marked with a breve, ᴗ ), except
    • If a short vowel is followed by two consonants (or double-consonants like x = ks), which means that the syllable is pronounced with a consonant at its end, that syllable is long.

So, if we were to apply this system to English, leg and rib would be treated as short, but hand and neck as long.

Turning to Latin, where we’ll stay for the rest of this piece, a pair of Latin words such as fās est (“it is right”) would sound – or ‘scan’ – as two long syllables (– –), pater (“father”) as two shorts (ᴗ ᴗ), and māius flet (“he weeps louder”) as long-long-short (– – ᴗ). Our life is made easier in Latin when we mark all naturally long vowels with the macron above them (e.g. leō), so we know that they are to be sounded differently. But the world’s a funny place, and most of us who work professionally with Latin don’t bother doing this.

J.H. Drake (ed.), Selected Fables of Phaedrus with indicated quantities (Allyn & Bacon, Boston, 1894).

With this as the basic system, let’s focus now on a single metre – the dactylic hexameter. This name sounds terribly technical, but it’s simple enough: six (hexa-) measures (metra) of finger-shaped (dactyli = Greek δάκτυλοι, daktuloi = “fingers”) rhythms. Take a look at your finger to reveal the beat: it’s composed of three parts, a long part first, then two short bits (ᴗ ᴗ). Repeat this unit (a ‘foot’ of the line as a whole) six times, and you have the basic form of the hexameter.

But since an endless stream of this ‘tum-ti-ti’ triple-rhythm would be mind-numbingly dull, the end of each line is signalled by replacing the last foot with its metrical equivalent of two longs (– – : the solemn ‘spondee’). In fact, you can mix up any of your dactyls with this spondee whenever and wherever you fancy, or when the words you need to use demand it. You then keep going for as long as you, or your audience, can stomach.

“Homer”, Jean-Baptiste Auguste Leloir, 1841 (Louvre Museum, Paris).

The rhythm of the dactylic hexameter is as old as the hills. It’s there from the very start of Greek literature, from the first to last line of Homer’s epic poems the Iliad and Odyssey (perhaps written in the 8th century BC). It also was the metre that marked a new start for Latin poetry. Quintus Ennius (239169 BC), a trilingual southern Italian who made his living as a Rome-sponsored poet, decided to reject the native Italian verse-forms (which operated on rules so different that scholars don’t fully understand them!) and to adopt instead the iconic rhythm of the Greeks’ dactylic hexameter.

The good news was that this placed his own poetry in the grand and unbroken tradition that already stretched back for almost a millennium. The bad news: Latin is much less suited to the metre, as the language has appreciably fewer short syllables than Greek. So you have to work harder to hammer out valid dactyls. Yet the die was cast, the mould set, and Ennius and his myriad successors had to make do.

Let’s look at the very first line of Latin hexameter poetry, the opening of Ennius’ mytho-historic epic, the Annales:

Mūsae, quae pedibus magnum pulsātis Olympum

Muses, you who strike great Mt Olympus with your feet!

Its six feet break down easily enough: Mūsae (– –), quae pedi– (ᴗ ᴗ), –bus mag– (– –), –num pul- (– –), –sātis Ol– (ᴗ ᴗ), –ym-pum (– –). That weighty opening word Musae sets the tone well. But it’s worth noting that Ennius couldn’t have begun the line with the proper Latin word for goddesses, deae (ᴗ), or even for Italy’s native Muses, Camēnae (ᴗ – –), because neither opens with the long syllable that must start every hexameter line. In the rule-bound world of Greek and Latin metre, either the verse scans properly, or else it does not, and therefore cannot be written at all: a nearly metrical line is an entirely impossible line. So stay sharp.

The Nine Muses: 3rd-cent. AD mosaic from Augusta Treverorum (modern Trier) (Rheinisches Landesmuseum Trier, Germany).

The hexameter poses some other problems. Imagine, as a good-ol’-fashioned Roman, you want to include in your poetry some concepts that you hold dear – vēritās (“truth”) and Latīnitās (“good Latin”), say, or being dīligens (“diligent”) or perspicax (“insightful”). Major problem, I’m afraid: none of these words can ever be used. Why? Because they contain the rhythm long-short-long (), which can’t fit anywhere in the line.

Even if you put these words in a different grammatical case (e.g. vērĭtātem or perspĭcācī), you can see you’re still stuck. So we can have the strange scenario that the most famous poem in Latin, Virgil’s Aeneid (20s BC), which is all about the founding of a new and world-changing state, cannot use the proper Latin word for that, cīvĭtās. And you can forget about including philosophia and its outlandish sequence of five shorts!

What about if your poem has been paid for by someone whose name just doesn’t fit into it? How to name-drop and flaunt your gratitude? Virgil (7019 BC) had this problem with his Eclogues (c. 38 BC), in which he really ought to mention his patron Pollio (). The only way to squeeze his name in is to cut off the final long “o” by ensuring that it is obscured, or “elided”, by a following word that begins with a vowel. If the mouth doesn’t close when saying these two words, the unwieldy –ō can hopefully be slurred out of the picture. So Virgil does exactly this: Polliō amat (“Pollio loves,” 3.84, 88), akwardly forging the dactyl Polli’ am– (ᴗ ᴗ).[2]

Don’t stop the flow: Eclogues 8–15: detail from the Palatine Virgil, copied c. 500 (Vat. Pal. Lat. 1631, 15v).

Ovid (43 BC c. AD 17/18), writing a generation later, is rather cheekier. In one of his poetic “Letters from the Black Sea” (Epistulae ex Ponto 4.12), he rebukes his friend Tuticanus (Tūticānus), who has impudently complained that no poem has been dedicated to him. The reasons, Ovid says, are clear enough: with such an awkwardly shaped friend, my hands are tied. I’m hardly going to chop your name up (nomen scindere) into two verses (in geminos versus), ending one line with Tūti­– and beginning the next with -cānus! Nor am I going to butcher it by mispronunciation, so that with an artificially short “a” it can somehow fit: Tūticănum. Your name, dear boy, just can’t cut the dactylic mustard. But don’t worry, Tutes, it’d work a treat in other metres – in iambics, in hendecasyllables, or even in galloping galliambics.[3] Perhaps more on those another time.

You may fairly wonder whether this system survive at all in modern poetry. In English, it’s a quite different story: the general plan is instead to use patterns of natural word stress (‘accent’), of rhyme, and often of both. Most of Shakespeare’s lines, for instance, are sequences of ten syllables, with the word stress on every even-numbered syllable. As Twelfth Night (1601?) famously opens,

If music be the food of love, play on!

Still, a few hardy souls have tried to match ancient Greek and Roman metres with the English word accent. Take an exchange from Henry Wadsworth Longfellow’s Courtship of Miles Standish (1858), and try to ride the hexametric wave of dactyls and spondees:

You are a writer, and I am a fighter, but here is a fellow
Who could both write and fight, and in both was equally skilful!

Romping stuff. Well, here’s more:

Yes, he was equally skilled, as you say, with his pen and his weapons…
Truly a wonderful man was Caius Julius Caesar!

Lord Tennyson, never one to mince his words, had other ideas about such endeavours:

These lame hexameters the strong-wing’d music of Homer!
When was a harsher sound ever heard, ye Muses, in England?[4]

Well, Longfellow or Wrongfellow, it was a very different story for the Greeks and Romans. Far from sounding repetitive or like trivial jingle-jangle, this was the mightiest metre of antiquity, big enough to convey epic tales, grand enough to channel divine hymns and powerful enough to pack the punch of satire.

Almost all the most famous poets of antiquity, not just Homer and Virgil, but the mighty names of Hesiod, Callimachus, Lucretius, Horace, Lucan, Juvenal and many scores more used the hexameter over countless thousands of lines. And Ovid, author of the poem that has influenced western art more than any other, the Metamorphoses (AD 8), chose this very verse form for his playful fifteen-book tour through each and every myth he could conjure up. Three thousand years on, and the hexameter remains a classic – a vintage staple that may not have rhyme but still has good reason to resonate.

David Butterfield is Senior Lecturer in Classics at the University of Cambridge and a Fellow of Queens’ College. He became interested in metre in 2003, on realising he had no clue how to get Aeneid 9.9 to scan.

Further Reading

There is no good introductory handbook to Greek and Latin metre. At the more technical level, there is plenty of useful information in the twin volumes of D.S. Raven (Greek Metre, London, 1962; Latin Metre, London, 1965; both reprinted by the much-missed Bristol Classical Press in 1998), but Martin West’s Greek Metre (Oxford UP, 1982; issued as a still substantial Introduction in 1987) remains magisterial.

A series of ten lectures on Greek and Latin metre, with accompanying handouts, can be found on Antigone here.


1 This hard-won accolade is shared with “squelched”, “scrunched”, “scratched” and “scrounged”, although none of these requires the speaker to squeeze eight consonants into one syllable.
2 The Latin and English text of this passage can most conveniently be read here.
3 This playful poem can be read in Latin here and English here.
4 For this and Tennyson’s other “Experiments in Quantity”, which expose the artificiality of such a practice, see here.