Horace’s own statements about the models for his odes are unequivocal: he portrays himself as a poetic craftsman working in the tradition of Greek lyric poetry as it was practiced about 600BC on the island of Lesbos by the Greek poets Alcaeus and Sappho (C.1.1:34, 1.26:9-12, 1.32:5, 3.30:13-14, 4.3.11-12, 4.6:35). Already in the “Iambi” he had used meters drawn from the poetry of Archilochus of Paros, a point underscored in E.1.19:23-25 where he claims that he “first displayed Parian iambi to Latium, following the meters and spirit of Archilochus, but not the subject matter or words that harried Lycambes.” Archilochus has many other meters besides iambic in his poetry, but the term came to be used of any “blame poetry” with a greater or lesser degree of hostility. As a matter of fact, all but two of the metrical combinations in Horace’s “Iambi” are attested in Archilochus, and the two missing systems might be attested if we had more of his poetry (Mankin, 15-16). It was common for Roman poets to ground their claims to originality in what they perceived as the vastly superior art of Greece, since they believed it was possible for them to be original only in relation to other Roman poets, so Horace began by writing his “Iambi” in the manner of Archilochus and then his “Carmina” in the manner of ancient Lesbos, by which he really means Alcaeus. While he professes himself the follower of Aeolic poetry, we find very little of Alcaeus’ spirit, content or emotional tone and none of Sappho’s in his odes. What he copied was their meter. Horace took his greatest pride in the technical accomplishment of adapting the meters of Lesbos to Latin versification. In the last ode of the three books published in 23BC, he states that “Not all of me will die, but a mighty part will escape the death-goddess Libitina: I shall continue to grow ever fresh with fame of after time, so long as the Pontiff climbs the stairs of the Capitol with the silent Vestal. I shall be declared, where wild Aufidus thunders and Daunus poor in water has always ruled his rustic people, as one risen high from humble origin who was the first to spin the meters of Aeolic song into Italian verse” (C.3.30:6-14). Fate has given him “a small domain and the fine breath of Greek song and a capacity to scorn the malicious crowd” (C.2.16:37-40). Because of Melpomene’s favor, the lush landscape around his Sabine farm “will make him famous for Aeolic song” (C.4.3:10-12).
The majority of the love poems are written in combinations of the five Asclepiadic meters, but the Aeolic forms Horace used most often were the four-line Alcaic and Sapphic stanzas. Of the 103 odes, 37 are Alcaic and 25 Sapphic (in addition to the “Carmen Saeculare”). Catullus had already written two poems in the Sapphic stanza (11 and 51), but their loose, rather linear organization provided no paradigm for the later poet. Horace didn’t merely adopt the two stanzas directly from Greek, he introduced a number of subtle metrical changes that enhanced their unity, gravity and structural effectiveness in Latin. His proud boast of being first “to spin the meters of Aeolic song into Italian verse” stands, and stood unchallenged in Latin poetry; no one else ever tried to follow the poetic trail he blazed. This is perhaps a good index of the overwhelming difficulties in modifying Greek metrical schemes to the Latin language. It is also, perhaps, an index of how radically unidiomatic–perhaps revolutionary–these rhythmic forms may have sounded to Roman ears.
The rhythms of both Greek and Latin poetry are based on the quantitative length of syllables, not on stress accent as are English rhythms. Meter is measurement, and what Latin meter measures are set patterns of syllables pronounced for a relatively long or short duration. Latin also had a very light stress accent on each word, probably mainly tonic, which fell on either the penult (second syllable from end) if it was long, or the antepenult (third syllable from end) if the penult was short. The stress accent sometimes coincided with long syllables and sometimes did not, so one might imagine it moving in a kind of counterpoint to the quantitative meter. But it had absolutely no effect on the quantitative meter. Poets could, however, occasionally play the stress accent off against the quantitative rhythm for certain local rhetorical effects. The stanzas Horace fashioned from his Greek exemplars possess an unrivaled sense of architectural solidity and cohesiveness. Their receptive metrical framework permitted him to organize words for the maximum of tension and relational impact. Yet these same stanzas never become stiff or inflexible, as if they were mere formal containers into which the poet poured his content. They progress from beginning to end of poem with an equally architectonic movement. Every stanza seems to dovetail effortlessly with every other stanza. Very often a single sentence will unfold over two or more stanzas, with a clear sense pause at the end of each to mark the stages. But even when the syntax of one stanza enjambs into the next, something that occurs only occasionally, the enjambment never produces any feeling of disorientation or carelessness. We always know precisely where we are in a Horatian ode and when, at the end, the final word falls into its predestined place, all movement ceases with a finality we feel like the completion of an overarching superflux.
Traditional English meter, by contrast with Greek and Latin, measures two rhythmic components: (1) the number of stresses and (2) their position within a fixed number of syllables. For this reason, it is called accentual-syllabic meter. The most common form is iambic pentameter or decasyllabic meter, where five stresses are nominally disposed over the even syllables of a 10 or 11-syllable line. Unlike Greek or Latin meter, however, accentual-syllabic meters permit the poet to vary the position of stresses with a great deal of freedom. The skillful positioning of stresses in unexpected positions or patterns produces both rhythmic variation and powerful expressive effects. This was the meter Spenser, Shakespeare, Milton, Wordsworth, Tennyson, Hardy, Yeats and Frost all employed for most of their poetry. It has been common in the long history of Horatian translations to use accentual-syllabic meters to render the quantitative meters of his odes. These meters are often deployed in rhymed couplets or other stanzaic forms–some traditional and some nonce inventions–that lack even the slightest organic relationship to the Latin stanzas. The adoption of such rhythmic strategies often trivializes Horace, making him sound like a glib 17th century Cavalier poet.
In translating Horace, however, “The choice of meter is the first and most vital consideration” as L. P. Wilkinson noted over 50 years ago in Horace and His Lyric Poetry (151). The following translations attempt to reproduce the internal stanzaic dynamics as well as the interstanzaic movement of the Horatian ode by using accentual templates for the quantitative rhythms. An accentual template is an array of stressed and unstressed syllables organized in approximately the same pattern as the long and short syllables of the original. Like a physical template, it provides a metrical model for the complete stanza. In the following templates, / = a stressed syllable, v = an unstressed syllable, x = either / or v and || = the principal break or pause (‘caesura’) where word-end almost always occurs. One of Horace’s most important departures from his Greek models was to fix the position of the caesura in each metrical form. The list of templates is divided into two parts: (1) the five Asclepiadic meters, with their official names, and then (2) the Sapphic and Alcaic stanzas.
Because English, like the other Germanic languages, cannot consistently produce runs of three contiguous stressed syllables due to the phenomenon of stress timing, those sections of the Horatian meters that regularly have three long syllables – – – are usually rendered by / v /. This is especially true of the following Asclepiadic meters, all of which open with three long syllables. For example, the Asclepiad in Horace is always
– – – u u – || – u u – u x
and never
– u – u u – || – u u – u x.
The practice of substituting / v / for – – – has long been common in German, which enjoys a magnificent tradition of accentual template poetry that extends back several centuries. Readers who know German might like to explore the works of Platen, Schiller, Goethe, Hölderlin and Mörike to name just the outstanding. In the second section on Horace’s Aeolic meters, I’ve included the quantitative scansions of his original Latin forms for comparison with the English accentual templates. Some concluding general remarks then explain the main differences between the Latin and English stanzas.
I. The Asclepiadic Meters
The components of the Asclepiadic meters are as follows:
- Asclepiad: / v / v v / || / v v / v x
- Greater Asclepiad: / v / v v / || / v v / || / v v / v x
- Glyconic: / v / v v / v x
- Pherecratean: / v / v v / x
1. The First Asclepiad:
/ v / v v / || / v v / v x.
This is a stichic meter, that is, it consists of Asclepiads repeated continuously in a series of lines. Examples: C.1.1; 3.30; 4.8.
2. The Second Asclepiad:
/ v / v v / || / v v / v x
/ v / v v / || / v v / v x
/ v / v v / || / v v / v x
/ v / v v / v x.
Three Asclepiads are followed by a Glyconic. Examples: C.1.6, 15, 24, 33; 2.12; 3.10, 16; 4.5, 12.
3. The Third Asclepiad:
/ v / v v / || / v v / v x
/ v / v v / || / v v / v x
/ v / v v / x
/ v / v v / v x.
Two Asclepiads followed by a Pherecratean and a Glyconic. Examples: C1.5, 14, 21, 23; 3.7, 13; 4.13.
4. The Fourth Asclepiad:
/ v / v v / v x
/ v / v v / || / v v / v x
/ v / v v / v x
/ v / v v / || / v v / v x.
This stanza consists of two distichs, each made up of a Glyconic and an Asclepiad. Examples: C.1.3, 13, 19, 36; 3.9, 15, 19, 24, 25, 28; 4.1, 3.
5. The Fifth Asclepiad:
/ v / v v / || / v v / || / v v / v x
This system is just a series of Greater Asclepiads written in lines. Examples: C1.11, 18; 4.10.
II. The Sapphic and Alcaic Stanzas
1. The Sapphic Stanza:
/ v / v / || v v / v / x
/ v / v / || v v / v / x
/ v / v / || v v / v / x
/ v v / x.
Examples: C.1.2, 10, 12, 20, 22, 25, 30, 32, 38; 2.2, 4, 6, 8, 10, 16; 3.8, 11, 14, 18, 20, 22, 27; 4.2, 6, 11; Carmen Saeculare.
Horace’s Latin Sapphic Stanza:
– u – – – || u u – u – x
– u – – – || u u – u – x
– u – – – || u u – u – x
– u u – x.
2. The Greater Sapphic Stanza:
/ v v / v / x
/ v / v / v v / || / v v / v / x
/ v v / v / x
/ v / v / v v / || / v v / v / x.
Example: 1.8.
Horace’s Latin Greater Sapphic Stanza:
– u u – u – –
– u – – – u u – || – u u – u – –
– u u – u – –
– u – – – u u – || – u u – u – -.
2. The Alcaic Stanza:
v / v / v || / v v / v x
v / v / v || / v v / v x
v / v / v / v / x
/ v v / v v / v / x.
Examples: There are too many to list, but the Alcaics in Book 1 are C.1.9, 16, 17, 26, 27, 29, 31, 34, 35, 37.
Horace’s Latin Alcaic Stanza:
x – u – – || – u u – u x
x – u – – || – u u – u x
x – u – – – u – x
– u u – u u – u – x.
General Remarks on the Aeolic Stanzas
1. In Alcaeus, the first syllable of the first three lines of the Alcaic stanza may be short. In Horace, it is normally long. In Alcaeus, the fifth syllable of the first three lines may be short. In Horace, it is always long.
2. In Sappho and Alcaeus, the fourth syllable of the first three lines of the Sapphic stanza may be short. In Horace, it is always long.
3. The cumulative effect of all these changes by Horace was to give the Aeolic meters a greater weight, dignity and almost legato tempo. The English accentual versions are actually rather closer to the Greek originals in their lighter movement.
4. Those who remain sceptical about the possibility of accentual templates to represent quantitative stanza forms should remember that the English stress accent is not unitary. It consists, in order of importance, of (1) tone, (2) duration and (3) force. In some research, duration or length of articulation appears even more important than tone. The component we traditionally take to be most prominent, force of articulation, is in fact the least prominent element. It is, therefore, quite feasible for English templates to approximate, if not equal, the aesthetic effects of the Latin meters.