I hope to start graduate school next fall in Classical Studies. Having worked on this undergraduate research topic now for a couple years, I intend to submit an expanded version of the paper as a writing sample to graduate schools in the hopes of continuing my research and developing it into a Master’s thesis.

“...with Alcaics it is no light task to reproduce the gathering wave of the first two lines, the thundering fall of the third and the rapid backwash of the fourth…” -L.P. Wilkinson, Horace and His Lyric Poetry (1945) p. 1521

This paper considers the way the Romans defined and explained the alcaic meter, the meter used most often in the Odes of Horace, the first century BCE master of Latin lyric, and how it can enhance our reading of his poems. A major aspect of the ancient description of meter was an emphasis on the ictus, also known as verse beat or pulse, which was an artificial accent, mentioned by Horace and Quintilian and used by the ancient grammarians, the teachers of the Latin language, to help students scan a line aloud.2 The ictus or verse beat was a way of measuring and dividing the line into feet or larger metrical pieces. The various ways these grammarians, the grammatici Latini, described Latin verse affected how it was read, and how it was read affects how we perceive it. As we read Latin poetry, the very local effects of sound in a specific line of a specific poem can be as important as the larger thematic issues raised by the poem itself.

“If we take the idea of a poetic language seriously, it can be defined first as a language in which the sound of the words is raised to an importance equal to that of their meaning, and also equal to the importance of grammar and syntax.”

- Kenneth Koch (1998) 20

The interplay of ictus and accent is already very familiar to modern students and scholars, not only of English poetry, but also of Latin poetry—from its pervasive and important effects in epic verse: it is commonly accepted that Latin epic generally aims for syncopation between accent and ictus in the middle of the line and coincidence at the end. Such interaction between natural word accent and artificial ictus, or verse beat, can also be useful in our responses to other meters, including lyric.

This paper will use Horace’s most well known poem, Odes I.9, the Soracte Ode, to illustrate the usefulness of ancient descriptions of meter, including accent and ictus, in a Latin lyric.

Latin verse is comprised of a pattern of long and short syllables. The basic pattern of the alcaic stanza is two lines of eleven syllables (hendecasyllabic), followed by one line of nine syllables (enneasyllabic), and a final line of ten syllables (decasyllabic).3

When studying Latin poetry, it is important to look at how the Romans described their own versification. The metrical composition of the individual lines of the alcaic stanza is variously explained by the ancient grammatici Latini.4 The grammatici Latini were ancient teachers of language who wrote about various aspects of Latin. Although most of the texts written by these grammatici come from the fourth and fifth centuries CE, they help us see things we would not see otherwise, even in texts from the Classical period (2nd century BCE – 2nd century CE), since they drew upon teachings from centuries earlier.

One prominent grammaticus of the 4th century CE, Marius Victorinus, gives a varied and thorough analysis of Horace’s meters. He divides the hendecasyllabic first two lines of the alcaic stanza into two halves (Keil VI. 166, 172, 268); the first half he describes as an iambic syzygy (x ? u ?) plus half a foot (?), a syzygy being the combination of two feet into one metrical unit. He describes the second half as two dactyls (? u u ? u u). This division puts a probable ictus (or verse beat) on the second, fourth, sixth, and ninth syllables. This acoustic feature is marked for the eye with a vertical line above the syllable.

Diagram of Latin verse patterns

Caesius Bassus, an earlier grammaticus of the first century CE divides the second hemistich (half line) differently (Keil VI.268): a trochee (? u) and an iamb (u ?) followed by another iamb.5 This is different from Victorinus’ division because it adds an ictus on the final syllable.

Diagram of Latin verse patterns

As mentioned above, the alcaic stanza has two lines of eleven syllables, followed by a nine-syllable third line and a ten-syllable fourth line. This enneasyllabic third line tends to feel slower because of the three consecutive long syllables in the middle. I will employ Marius Victorinus’ descriptions because they are the clearest despite their complicated terminology. He describes the enneasyllabic line in terms of the iambic alternation of short and long syllables, calling it by the cumbersome name hypercatalectic iambic dimeter (Keil VI.1670). Hypercatalectic in this instance means a syllable has been added on to the end of the metrical unit, and dimeter means two pairs of iambic feet. The ictus would measure the line into feet as shown below.

Diagram of Latin verse patterns

He then calls the decasyllabic fourth line a hypercatalectic dactylic trimeter, hypercatalectic again meaning an extra syllable is added on to the end of the metrical unit, and a dactylic trimeter being just three dactylic feet:

Diagram of Latin verse patterns

Upon close inspection, this example is not unproblematic since Victorinus’ division of the decasyllabic final line has it ending in an iamb (u?) though the last two syllables are actually a long syllable first and then an anceps (a short or long syllable treated as long at the end of the line). One way to explain this is to presume a metathesis, or transposition, of the last two syllables, which is called anaclasis in the Grammatici Latini, usually referring to a long syllable switching places with a neighboring short syllable (Halpborn, James, Martin Ostwald and Thomas Rosenmeyer, The Meters of Greek and Latin Poetry (1980) p. 121). The variations and descriptions presented here are a prime example of some of the difficulties in dealing with the ancient evidence of the grammatici Latini. They are often contradictory, difficult to interpret, and sometimes have mistakes. Nevertheless, it provides us with a window to the way meter was taught in the first few centuries CE, and the way meter is taught cannot help but have some influence in the way meter is understood.

To return to the alcaic, Victorinus provides other alternatives. For example, as you can see below, he calls the decasyllabic fourth line two dactyls (? u u) joined with two trochees (? u) (VI.111 Keil). This puts an ictus on the first, fourth, seventh, and now also on the ninth syllables.

Diagram of Latin verse patterns

Marius Victorinus’ third option for this line is what is called an archebulean with a foot and a half removed from the beginning (Keil VI.126 and VI.269). Below is the pattern of an archebulean line, and how it can be modified to arrive at the pattern of the decasyllabic line.

Diagram of Latin verse patterns

As the examples seem to suggest, one of the preoccupations of the grammatici Latini is the interchangeability of the different meters. Marius Victorinus and Caesius Bassus, though writing two hundred years apart, both tend to explain the lines of the alcaic stanza as made up of iambs and dactyls (Keil VI.166.10- 167.2 and VI.268.10-269.15). Why do they use dactylic and iambic meters to explain alcaics? One explanation is that these iambic meters were more familiar to the audience, therefore easier to use as a basis for describing a less common meter. Another could be the derivation theory of meter: it derives all metrical patterns from iambs and dactyls.6 Nevertheless, the different divisions do not result in significant differences in the position of the ictus.

The first nine poems of Book I of Horace’s Odes are called the Parade Odes: each one was written in a different meter, demonstrating Horace’s metrical mastery. The last of these Parade Odes, Odes I.9 serves as the model of an alcaic. Though it is in Latin, we are not concerned here with the sense, just the sound of the accents and beats.


Vides ut alta stet niue candidum Soracte nec iam sustineant onus siluae laborantes geluque flumina constiterint acuto?

See how Soracte under thick-fallen snow Stands up resplendent, nor can the Woods any longer bear the load, and Rivers have flowed to a halt in sharp ice.

dissolue frigus ligna super foco large reponens atque benignius deprome quadrimum Sabina, o Thaliarche, merum diota.

Thaw out the cold by piling the fire up With plenty of logs and pouring lavishly From its Sabine two-handled pitcher, O Thaliarchus, the four-year-old wine.

permitte diuis cetera, qui simul strauere uentos aequore feruido deproeliantis, nec cupressi nec ueteres agitantur orni.

Leave all the rest to Gods, for as soon as they Have laid the storm-winds battling it out upon The seething ocean, then are neither Cypresses shaken nor ancient rowans.

Quid sit futurum cras, fuge quaerere, et quem fors dierum cumque dabit, lucro adpone nec dulcis amores sperne, puer, neque tu choreas,

Beware of asking what may tomorrow bring, And enter up as profit whatever day Good Luck may grant you, nor in boyhood Say No to sweet love-affairs and dances,

donec uirenti canities abest morosa. Nunc et Campus et areae lenesque sub noctem susurri composita repetantur hora

While youthful green is free from cantankerous White hair. So now let Campus and public square And gentle whispers after nightfall Often be sought at an hour agreed on.

nunc et latentis proditor intumo gratus puellae risus ab angulo pignusque dereptum lacertis aut digito male pertinaci.

Sweet also now, betraying her hiding-place, Is girlish laughter heard from a secret nook And keepsake snatched away from upper Arm or a seeming-reluctant finger.

* The Latin text is that of Shackleton Bailey, Q. Horati Flacci Opera (1995) * The translation is that of Lee, Horace Odes and Carmen Saeculare (1998) p.17

In light of the treatment of the meter by the grammatici Latini, it is interesting to consider the relation of natural word accent to the metrical ictus. The rest of this paper will focus particularly on the pivotal third line, what Wilkinson called above the “thundering fall” of the stanza.

The third line of the stanza is considered by scholars to be the fulcrum, the spot where the modern poet Rosanna Warren says the stanza “seems to lose its footing mid-stride.”7 As mentioned above, with its three consecutive long syllables in the middle, the third line of the alcaic stanza is a slower line, especially when contrasted with the quicker dactylic fourth line. However, the feeling of slowing down cannot come just from the arrangement of long and short syllables since the first seven syllables of this third line are exactly like those of the first two lines. Instead, this slower feel comes from the common distribution of words, with their accents, that Horace uses in these lines.

Each of the six enneasyllabic lines in Odes I.9 ends in a three syllable word: gelúque (3), Sabína (7), cupréssi (11), amóres (15), susúrri (19), lacértis (23). Having a three-syllable word at the end of the enneasyllabic third line of the stanza allows the accent and ictus to coincide as the line comes to an end. This coincidence between accent and ictus emphasizes the metrical pattern as the line comes to a close. Here’s an example (with natural word accent marked with capital letters):

Diagram of Latin verse patterns

Ending the nine-syllable line in a three-syllable word creates a break after the long syllables in the middle of the line.

Diagram of Latin verse patterns

If a single word occupies those three long syllables in the middle (a molossus), it can emphasize the slow feel of the line. When it falls in a metrical position like this, with an ictus on the first and third syllable and a word accent on the second, it is known as a swell.8 In Horace’s alcaic Odes, this trisyllabic molossus swell is the most common pattern;9 it appears in three of the six nine-syllable lines in Odes I.9 above: in line 7 (quadrímum), 23 (deréptum), and also 19 (sub nóctem, since the preposition and the noun it governs would be treated prosodically as one word for the purpose of accent). The accent falls between the verse beats even in lines that do not have the molossus swell, like line 3 (sílvae laborántes gelúque) and line 15 (adpóne nec dúlcis amóres).

The awareness of the way this meter was understood in ancient times can help us to hear and understand the rhythm of Horace’s Latin lyric. These prosodic features of the alcaic stanza become more apparent and significant when seen in light of the grammatici Latini and the relation of accent to ictus. Using the evidence of the grammatici Latini, we learn that Horace used syncopation in the middle of this pivotal line and coincidence at the end. This is important because, although sound does not have to be correlated with meaning, the poetic importance of those and similar features can be useful for better understanding Horace’s versification.


It is helpful and interesting to look at some English comparanda such as John Hollander and WH Auden. Hollander cleverly describes the alcaic pattern with an English alcaic stanza (11, 11, 9, and 10 syllable lines and 5, 5, 4, and 4 or 3 stresses, respectively).

Thís tíght alcáic stánza begíns with á
Mátched páir of lónger línes that are fóllowed by
Two shórter ónes, indénted thís way,
Máking the méter declaím in Énglish.

While Hollander’s stanza is a pedagogical tool, a stanza from WH Auden’s “In Memory of Sigmund Freud” is an example of an alcaic stanza in practice:

Ónly Háte was háppy, hóping to augmént
his práctice nów, and his díngy clíentéle
who thínk they cán be cúred by kílling
and cóvering the gárden with áshes.

The endings of each line appropriately mimic the feel of the Latin alcaics from which they are adapted. The 11-syllable first and second lines have iambic endings. In English this is an unstressed syllable then a stressed syllable, in Latin a short then a long. The 9-syllable line and the 10-syllable line have trochaic endings, in English a stressed then an unstressed, in Latin a long then an anceps. This Latin-inspired pattern is evident throughout the poems.


1Cf . Rosanna Warren, Fables of the Self: Studies in Lyric Poetry (2008) p. 31 “Let us hear now, in Alcaeus’ own music, the song from his broken world, in a meter which seems to lose its footing mid-stride, but recovers its balance, and heals in the order of the ear what on earth and in politics remains in ruins.” and David West, Horace Odes II: Vatis amici (1998) p. xx “…the drag of the third line and the acceleration of the fourth.”
2 For reference to ictus see Becker (2004)
3 Like all Latin quantitative meters, Horatian alcaics adapt Greek models. In positions one and five in the first two lines, Greek alcaics maintain an anceps, which is a variable syllable that can either be a long or a short. This variability emphasizes the iambic feel that will be discussed later. In taking the Greek model and making it their own, Latin poets made the fifth syllable always long, and the first syllable usually so. For more information on Latin meter see Raven, Boldrini, Halporn- Ostwald-Rosenmeyer, Nisbet and Hubbard, Drexler, and Crusius-Rubenbauer.
4 The surviving texts are collected in Keil’s eight volumes.
5 This series of long and short syllables corresponds to an asclepiadean ending (? ? ? ? ? ?).
6 For more on the derivation theory, cf. Leonhardt, ‘Die beiden metrischen Systeme des Altertums’ (1989) who explains the Alexandrian view that all meters were derived from dactylic and iambic meters mutatis mutandis.
7 Warren, Fables of the Self: Studies in Lyric Poetry (2008) p. 31
8 Nussbaum, Vergil’s Metre (1986) p. 41
9 Nisbet and Hubbard, A Commentary on Horace Odes, Book I (1970) and Zinn, Der Wortakzent in den lyrischen Versen des Horaz (1940)


Allen, W. S. “Accent and Rhythm.” Prosodic Features of Latin and Greek: A Study in Theory and Reconstruction. Cambridge: Cambridge University Press, 1973.

Auden, WH. Collected Poems. Ed. E. Mendelson. Vintage International, 1976.

Aviram, Amittai F. Telling Rhythm: Body and Meaning in Poetry. Michigan, 1994.

Becker, Andrew S. “Non oculis sed auribus: The Ancient Schoolroom and Learning to Hear the Latin Hexameter.” Classical Journal 99: 313-22, 2004.

Boldrini, Sandro. Fondamenti di prosodia e metrica latina. Carocci, 2004.

Boldrini, Sandro. La prosodia e la metrica dei romani. Carocci, 1992.

Crusius-Rubenbauer. Romische Metrik Georg Olms, 1967.

Drexler, Hans. Einfuhrung In Die Romische Metrik. Wissenschlaftliche Buchgesellschaft, 1967.

Edmunds, Lowell. From a Sabine Jar. University North Carolina Press, 1992.

Halpborn, James, Martin Ostwald and Thomas Rosenmeyer.The Meters of Greek and Latin Poetry. Univ. of Oklahoma Press, 1980.

Hollander, John. Rhyme’s Reason: A Guide to English Verse. Yale University Press, 1981.

Kaster, R. A. Guardians of Language: The Grammarian and Society in Late Antiquity. Berkeley, Los Angeles, and London: University of California Press, 1988.

Keil, Heinrich. Grammatici Latini. Georg Olms, 1855-1880.

Lee, Guy. Horace Odes and Carmen Saeculare. Leeds: Francis Cairns, 1998.

Leonhardt, Jürgen. ‘Die beiden metrischen Systeme des Altertums.’ Hermes117 (1989): 43-62

Nisbet, R.G.M and Margaret Hubbard. A Commentary on Horace Odes, Book I. Oxford: Clarendon Press, 1970.

Nussbaum, G. B. Vergil’s Metre. Bristol: Bristol Classical Press, 1986.

Raven, DS. Latin Metre. Briston Classical Press, 1965.

Shackleton Bailey, D.R. (3rd edition). Q. Horati Flacci Opera. Stuttgart: Teubner, 1995.

Warren, Rosanna. Fables of the Self: Studies in Lyric Poetry. WW Norton and Company, 2008.

West, David. Horace Odes II: Vatis amici. Oxford University Press, 1998.

Wilkinson, L.P. Horace and His Lyric Poetry. Cambridge University Press (1945): 152.

Zinn, Ernst. Der Wortakzent in den lyrischen Versen des Horaz. Georg Olms, 1940.


Photo of Alexandria DeSio   Alexandria DeSio is a graduating senior from Chesapeake, Virginia. She is aclassical studies major who aspires to be a published collegiate professor. Inorder to reach her goal, she hopes to continue on to graduate school this fall. Alexandria appeared in last year’s inaugural issue of Philologia in a featured article by Audra Vasiliauskas. She has presented her research at various confer-ences around the country. Alexandria would like to recognize her family for being so steadfastly supportive of her studies. She would also like to thank Dr.Andrew Becker (foreign languages and literatures) for his incessant reassurance, guidance and encouragement. It is because of him that she discoveredher passion for Latin and her career path as a professor.