Horace's Victory Odes

HORACE'S 'VICTORY ODES': ARTIFICES OF PRAISE*

ARMAND D'ANGOUR

Horace's claim to engage with Pindar as a Greek lyric predecessor is implicit in hisprogrammatic introduction, the first Ode of Book 1.1 The first cursus vitae to be described,if eventually dismissed by Horace in favour of poetic stardom, is that of the most exaltedobject ofPindaric epinikion, the Olympic charioteer:

sunt quos curriculo pulverem Olympicumcollegisse iuvat, metaquefervidisevitata rotis palmaque nobilisterrarum dominos evehit ad deos. 2

there are some whose delight is to have gathered Olympic dustwith their chariot, and grazing the turning-postwith burning wheels and winning the victory-palmmakes them lords on earth and raises them to the gods.

The priamel, famously used in the opening verses of Pindar's Olympian 1, may be heardas a specifically Pindaric signature-device. The close echo of another fragment emphasizesHorace's declaration ofhis Pindaric credentials:

QEAAOTT68wv IlEV TlV' El)(ppatvowLV 'lTTTTWVTLlla\. Ka\. aTE</:>avOLTOUe;; 8' EV TToAuxpuaOLc; 6aAaIlOLC; ~LOTci'

TEPTTETaL 8E Kat TLC; ETT' 01811' (iALOVva\. 60q: t8LaaTEl~wv. 3

• I am grateful to the organizers of the Epinician Conference and Chris Carey in particular for theinvitation to think about this subject, and to Richard Rawles and Colin Sydenham for reading and commenting on this paper. The first half of my title had been proposed by Alessandro Barchiesi (the adjuncttitle echoing Putnam 1986 is mine), whom I was asked at short notice to replace; Barchiesi (2002, 2009)has illuminated the relationship of Horace and Pindar by showing how the Carmen Saecu/are is aPindaric-style paean. I am also indebted to Philip Hills for letting me see his prize-winning Cambridgedoctoral thesis on Horace's Fourth Book ofOdes (Hills 2000).

I Cf West 1995, p. 6. I prefer to speak of Horace's 'engagement' with rather than 'debt' to Pindar. AsFeeney (2009, pp. 207-08) has emphasized, Horace's reading of Pindar was inevitably mediated byHellenistic scholarship: 'at a stylistic level at least, homage to Callimachus may look oddly likehomage to Pindar'. Pindaric references, elements and parallels in Horace's Odes are well documentedin commentaries such as Nisbet and Hubbard (1970, 1978), and Nisbet and Rudd (2004). Studiesinclude Rummel 1892 and Harms 1936; and there are excellent detailed discussions and observationsin Fraenkel 1957, Davis 1991, Putnam 1986, and most recently Race 20 IO.

2 Odes 1.1.3-6. At line 32 Horace implicitly reaffirms his emulation of Pindaric choral lyric byreferring to tibiae, the counterpart of au/oi associated with performances of epinikia.

3 Pindar fro 221; the metre is dactylo-epitrite, as used in a large proportion of Pindar's epinikia.

57

58 RECEIVING THE KaMaS

One man takes delight in honours and crowns won by horseswith storm-swift hooves, some love living in gilded chambers,

while another man takes pleasure in crossing the sea-swell in a swift ship.

In the first three books of the Odes, a wide variety of similar allusions to the content and tex

ture of Pindaric epinikia are to be found." Odes 1.7begins with a priamel (laudabunt alii ...), andwhile not related to a specific epinikion, the Ode unmistakeably reproduces 'the wide sweep and

deliberately casual tmnsitions of the Pindaric encomium'. 5 Odes 1.12 famously takes its opening

motto quem virum aut heroa from Olympian 2 (though with a not insignificant reversal of the

positions of8E6v and av5pa).6 Odes 3.1-6, the 'Roman Odes', are notably Pindaric in their solemn

gmndeur and moralizing content.' 3.4 in particular, Descende caelo, reproduces a number of

specific Pindaric themes found in Pythians 1 and 8: the special status granted to poets, the linkbetween poetic and political harmony, and the myth ofGiants who represent mindless violence. 8

The Alcaic metre ofthe 'Roman Odes' is a reminder ofHorace's explicit self-identification in

his Epistles as Alcaeus, the singer of elite sympotic songs that are both personal in tone andpolitically engaged," But that identification, though predominant, is never all-encompassing, and

Horace's adoption ofa public persona and his claim to authority as a vates lift his Odes beyond the

associations of the Asclepiadic metres that he masterfully adapts and regularises. In the final Odeof the third book, even as he vaunts his priority in applying Sapphic and Alcaic rhythms to Latin

verse (3.30.13-14, princeps Aeolium carmen ad Italos / deduxisse modos), he proceeds to

introduce a subtle reminder ofhis Pindaric status. Inthe last lines ofthe Ode he turns to address theMuse, and bids her crown him with Delphic laurel, the emblem of Apollo, divine patron of the

Pythian festival:

sume superbiamquaesitam meritis et mihi Delphicamlauro cinge volens, Melpomene, comam.l"

take sought-after pride,Melpomene, in well-deserved praise,and with Delphic laurel graciously crown my hair.

4 As often noted, Pindar himself does not use the term epinikion, an Alexandrian subdivision ofenkiimion (the poet's own term); but it is useful to use the term to categorize the extant Odes inpraise of (in nearly all cases) athletic victors.

5 Nisbet and Hubbard 1970, p. 93.

6 Apart from the motto, there is little that is Pindaric in the Ode; as Feeney notes (2009, p. 221 n.47)'the most fundamental categories are not the same ... by the end of the poem it is by no meanscertain to which category Augustus belongs'.

7 Fraenkel 1957, pp. 260f. Griffiths 2002 makes the case for the Roman Odes being read as 'a singlegreat canto' (p. 78), outlining a 'pseudo-triadic' structure that would recall Pindar.

8 See Fraenkel 1957, pp. 276f. Nisbet and Rudd 2004 ad loc. note that the autobiographical first halfof the poem owes nothing to Pindar, and judge the Ode's rhetoric 'very un-Pindaric'.

9 Hor. Epist. 2.2.99: discedo Alcaeus puncto illius, cf 1.19.32-3; Woodman 2002, pp. 56-58.

10 Odes 3.30.14-16.

ARMAND D'ANGOUR: HORACE'S 'VICTORY ODES' 59

By ending the collection of Odes 1-3 with eomam, Horace draws attention to apotentially loftier self-identification, since 'hair' is the final word of the last Ode in thecollection of Pindar's Olympian Odes, written for one Asopikhos of Orkhomenos who 'inthe renowned vale of Pisa at the fame-giving games won crowns with winged wreathswith which to deck his youthful hair':

VfavKOATTOLe; rrup' ElJOOeOLe; THeme;EUTECI)(ivWUE KVOLliwV M6AWV rrrepoiot Xa{TaV. II

Despite Pindaric associations of this kind, however, no Ode in the first three-bookcollection presents itself as a praise poem comparable to any of Pindar's epinikia. Moreover,some of the Odes imply that there may be something suspect about overt personal praise fora living dedicatee. In Odes 2.4, for instance, 'Horace' feels the need to insist with earnesthumour that his praise of Phyllis is disinterested and beyond suspicion (22, integer laudo,fuge suspieari). 12 In 1.13, Lydia's praise of Telephus evokes violent physical symptoms ofjealousy:

cum tu, Lydia, Telephieervieem roseam, eerea Telephi

laudas braeehia, vae meumfervens difjieili bile tumet ieeur.

tum nee mens mihi nee coloreerta sede manent, umor et in genas

furtim labitur, arguensquam lentis penitus maeerer ignibus. 13

When you praise Telephus'srosy neck, Lydia, Telephus's

waxen arms, how my liver simmerswith indigestible bile.

Then neither my mind nor complexionstays fixed in place, moisture trickles furtively

on my cheeks, showinghow I am wasted to the marrow by slow-burning fires.

Here 'Horace' disintegrates in the face of praise for a rival in love." But could the poethimself ever remain truly integer in the process of offering overt praise to a living individual,without creating the suspicion of insincere flattery, distasteful abjection, or unacceptable irony?

II Pindar 0. 14.23-4; Oliensis 1998, p. 104, n.3.

12 Cf integer vitae in Odes 1.22.1, used as ironic self-praise by Horace for his poetic devotion toLalage.

13 Odes 1.13.1-8.

14 The humorous extravagance of the symptoms evokes an intertextual relationship with Sappho fr.31 and Catullus 51, and the context recalls Odes 1.8, in which Lydia has lured the athlete Sybaris,champion of discus and javelin, away from the campus.

60 RECEIVING THE KOMOS

The necessary exception to Horace's insistence on his disinclination to praise was inrelation to the Princeps. Horace's refusal to expand on Augustus's military and politicalexploits, as in the Callimachean recusationes of epic of Odes 1.6 (scriberis Vario) and oflyric in 2.12 tnolis longaferae), had a long and respectable literary pedigree. But Augustus'laudability is established in Odes 1.12, where he is given a proud place at the end of theparade of gods and heroes. IS After Horace's status as a public poet, destined for famebeyond his lifetime, had been confirmed by his composition of the Roman Odes and by hisringing claims in the coda to the first three books (3.30, exegi monumentum aereperennius...non omnis moriar etc.), the pressures to celebrate the regime of Augustus couldonly intensify. Having toyed with various elements of Pindaric style throughout his firstthree books, in June of 17 BC Horace triumphantly fulfilled a public commission with thecomposition and production of the Carmen Saeculare - the only Horatian Ode we know tohave been performed and, like Pindar's epinikia, performed chorally." Suetonius bracketsthe commission of the song with that of the composition of the fourth book ofOdes:

scripta quidem eius usque adeo probavit mansuraque perpetuo opinatus est, ut nonmodo saeculare carmen componendum iniunxerit sed et Vindelicam victoriamTiberii Drusique privignorum suorum, eumque coegerit propter hoc tribuscarminum libris ex longo intervallo quartum addere. 17

[Augustus] was so keen on his writings and thought they would last for ever that henot only ordered the composition of the Carmen Saeculare but also a celebration ofthe victory of his stepsons Tiberius and Drusus over the Vindelici [15 BC], and hepressured Horace on this account to add, after a long interval of time, a fourth bookto his three books of Odes.

Accordingly, Book 4 is where 'Victory Odes' are generally sought and found; and in thisbook Horace finally appears to give serious consideration to donning the wreath of theepinician poet. 18 Odes 4 and 14 show an obvious indebtedness to Pindaric epinikia; Odes 8and 9 embrace strongly Pindaric motifs; and Odes 5 and 15, while displaying other genericaffiliations, include strongly Pindarizing encomia of the Princeps and his achievements. Formany commentators, the generic identity of Odes 4 and 14 in particular is unquestioned:

15 However, the beginning and ending of the Ode, with its praise of and prayer to Jupiter, 'enablesthe poet to touch upon the ruler's greatness in dignified terms without eulogizing him directly'(Fraenkel 1957, p. 296). One wonders whether the Ode immediately following, 1.13, may be felt toundercut the expression of overly laudatory sentiments; cf Lync's observations (1995, pp. 207f.) ofthe way Horace 'takes back with one hand what he has given with another' (referring in particular tothe 'sapping' of Odes in Book 3 by those in Book 4).

16 Barchiesi 2002 argues that the Carmen is a 'neo-Pindaric paean'. A quasi-triadic structure haslong been suggested (e.g. Fraenkel 1957, p. 371), but Schmidt's performative reconstruction (2009,pp. 133-38) indicates that 'Horace keeps himself free of any schematism' (p. 138). Lyons 2010,pp. 41f., following in the footsteps of Wille 1967, vigorously reasserts the extent to which musicalperformance was a feature of Roman experience in Horace's time.

17 Suet. Hor. 20-5.

18 Cf Highbarger 1935.


'The epinikion for Drusus, placed before an ode addressed to Augustus, has its counterpartin the epinikion for Tiberius (4.14), again followed by an ode to the Princeps'. 19

Horace's encomia were bound to reflect, at least obliquely, praise of Augustushimself, even if other laudandi were explicitly in the frame. Thus while Drusus was theintended honorand of Odes 4.4, duly celebrated for his conquest of the Vindelici andRaeti, a decorous tribute to the emperor was no less warranted:

sensere quid mens rite quid indolesnutrita faustis sub penetralibusposset, quid Augusti paternusin pueros animus Nerones/"

They felt what a mind, what a nature dulynurtured within a favoured household,what the fatherly spirit of Augustustowards the Nerones could realise.

Odes 4.14, in honour of the young Tiberius, goes much further, beginning with thepoet's direct and fulsome address to the Princeps:

Quae cura patrum quaeve Quiritiumplenis honorum muneribus tuas,Auguste, virtutes in aevumper titulos memoresque fastus

aeternet, 0, qua sol habitabilisillustrat oras, maxime principum? ..21

What attention of senators or of citizensGiving full reward for honours due to you,through eulogies and mindful records, Augustuscan immortalise your virtues

For time to come wherever the sun surveysthe inhabited earth, 0 greatest ofprinces?

The Pindaric elements in both these Odes, affirmed not least by the long simile thatintroduces Odes 4.4, have been well rehearsed by commentators.F But in their content,tone, and manner, these poems also show significant divergences from Pindaric epinikia.First, even as Horace extols others, his eyes are on Augustus: and the blurring of the focusof praise for an individual victor, with the poet's gaze fixed on the potentate standing

19 Fraenkel 1957, p. 431.

20 Odes 4.4.25-8.

21 Odes 4.14.1-6.

22 E.g. Fraenkel 1957, pp. 426f.; cf Putnam 1986, pp. 81f., 236f.: 'to begin a lyric with an extendedsimile is a Pindaric gesture' (p. 85).


behind the ostensible honorands, is more Hellenistic than Pindaric." Closely connected tothis point is the fact that these Odes are not, of course, celebrations of athletic victories,but of successful military campaigns which would properly merit collective rather thanindividual praise." What is striking above all, perhaps, is the way Horace dwells on thethoughts or characteristics of the vanquished. Thus in Odes 4.4 it is Hannibal who has thebest lines, and those most laudatory of Claudian prowess. Indeed, the conflict is viewedfrom the defeated Carthaginian's perspective:

Carthagini iam non ego nuntiosmittam superbos: occidit, occiditspes omnis et fortuna nostrinominis Hasdrubale interempto.

nil Claudiae non perficient manus,quas et benigno numine Iuppiterdefenditetcuraesagacesexpediunt per acuta belli.25

Now I shall not send proud messengersto Carthage; perished, perishedis the whole hope and fortune ofour name, now Hasdrubal has fallen.

Nothing cannot be achieved by Claudian hands,which Jupiter with gracious powerdefends, and which wise attentionmanoeuvres through war's hazards.

Odes 4.14 does not repeat this imaginative focalization, but revels in listing the defeatedtribes and peoples with tellingly Homeric epithets <legis expertes Latinae / Vindelici (7-8),immanisque Raetos (14), Cantaber non ante domabilis (41), profugus Scythes (42), remotis... Britannis (45-6) - through to the final stanza:

te non paventis funera Galliaeduraeque tel/us audit Hiberiae,te caede gaudentes Sygambricompositis venerantur armis."

23 Pindar brings praise of others (family, trainers et at.) into his Odes, but (without resorting toBundyan extremes) the central focus on the laudandus is inescapable.

24 While sport requires the continued existence of competitors, military victory for one party entails thedefeat and perhaps even the obliteration of the other. Pindaric epinikia (0. I2, N.2, etc.) occasionallyallude to military engagements such as Himera and Salamis, but these are subsidiary to the praise ofathletic victories.

25 Odes 4.69-76.

26 Odes 4.14.49-52.


You the land of Gaul, which fears not death,heeds, and of rugged Hiberia,you the Sygambri who delight in slaughterworship, with their weapons laid down.

In view of the rich characterization of the defeated parties, Odes 4.4 and 4.14 mightas well be described as 'Defeat Odes' rather than Victory Odes. In this respect, they offera stark contrast to Pindar's epinikia, which never engage with the viewpoint of thevanquished. In one seeming exception, a passage from Pythian 8, the description of the(unnamed) losers is simply a schematic reversal of the situation of the winners - theirhappy homecoming, public acclaim and so on:

To them no joyful homecoming like yourshas the Pythian festival afforded.On their return to their mothers no happy laughteris raised about them; keeping clear of their enemies, they slinkdown alleyways, bitten by failure. 27

Even when he Pindarizes most fully, then, Horace diverges in significant ways fromPindaric epinikia. In fact, had Horace chosen to transform himself into a composer ofPindaric-style Victory Odes, he could have issued a clear statement of intent. One mightimagine him beginning his fourth book, for instance, with an imitation of the mostunmistakeably Pindaric of mottoes:

'optimum' dicamus aquam, sed ignesnocte praefulgent; neque, si deorumseu virum laudes celebrantur, absum -

Pindarus alter. 28

Water is best, we may say, but fireshines out in the night; nor, if deeds ofgods or men are being praised, do I absent myself

I, a second Pindar.

But this is not Horace. While keen to promote himself as lyricus vates par excellence,and despite the use of Pindaric structures, themes, mottoes, and phrases so studiouslyannotated by commentators, Horace never makes such a self-identification. Indeed, theRoman poet who comes closest to emulating Pindar must be the one to claim that such athing should not be attempted. Horace's apparently straightforward view of the matter isoffered in the heavily Pindarizing Odes 4.2:

27 Pindar P.8.83-7. Cf fr. 229 'those who lose are bound with a gag of silence, unable to face theirfriends'. In 0.8.69, the victor projects on to the losers 'a most hateful homecoming, words lesshonouring, an obscure path'.

28 Cf Horace's designation of Ennius as alter Homerus (Epist. 2.1.50); for the use of Sapphics here,and the late caesura in the first line, see my observations on metre in the last section of this article.


Pindarum quisquis studet aemulari,lulle, ceratis ope Daedaleanititur pennis, vitreo daturus

nomine ponto.

monte decurrens velut amnis, imbresquem super notas aluere ripas,fervet immensusque ruit profundo

Pindarus ore,

laurea donandus Apollinari,seu per audaces nova dithyrambosverba devolvit numerisque fertur

lege solutis ...

multa Dircaeum levat aura cycnum,tendit, Antoni, quotiens in altosnubium tractus: ego apis Matinae

more modoque,

grata carpentis thyma per laboremplurimum, circa nemus uvidiqueTiburis ripas operosa parvus

carmina jingo. 29

Whoever strains to emulate Pindar,Julius, relies on wings of waxit la Daedalus, and is destined to lend his name

to the glassy sea.

Like a stream cascading down a mountainwhen rains have nourished it over its familiar banks,Pindar seethes and unrestrained rushes

with his fathomless voice,

worthy of the reward of Apollo's laurel,whether he whirls down new words in bold dithyrambsand is borne along on rhythms

free from restraint ...

Many a breeze lifts the swan of Dircewhenever, Antonius, he heads towardsthe lofty reaches of the clouds; J, in the mode and manner

of a Matine bee

culling tasty thyme with much toilaround the grove and banks of watery Tiburin my small way sculpt my songs,

the product of much effort.

29 Odes 4.2.1-12, 25-32.


The Ode is addressed to Iullus Antonius, son of Mark Antony by Augustus' sisterFulvia, a prominent nobilis and a favourite of Augustus. The significance of Horace'ssuggestion (33-4, concines maiore poeta plectro / Caesarem ...) that Iullus is well placedto sing Augustus' praise in Pindaric style has occasioned much debate.30 But the image ofmaius plectrum requires us first and foremost to imagine a larger and louder instrumentthan the lyra, i.e. the cithara, an instrument associated with public performance, whetherof lyric or epic. Whereas the freedman's son could not publicly have presented his ownrelationship to the Princeps as one of social equivalence, of amicitia on the level ofPindaric philia or xenia, Iullus could have done so; and he might also have been expectedto playa prominent part in the celebrations greeting Augustus' victorious return. Horaceimagines Iullus walking alongside Augustus in the procession, while the poet presentshimself as a voice in the crowd chanting rhythmical tributes to the victorious Princeps;"

tuque dum procedis 'io Triumphe'non semel dicemus, 'io Triumphe'civitas omnis dabimusque divis

tura benignis.t'

And as you lead the way, 'Hail, god of Triumph'again and again, 'Hail, god of Triumph'we citizens all will proclaim, and we will offer incense

to the favouring gods.

Horace's reasons for rejecting close Pindaric identification were therefore a matterboth of social propriety and aesthetic choice, issues connected to an even greater extent inAugustan Rome than they had been for Pindar and his Greek successors in panegyricverse.r" But the notion of his composing 'Victory Odes' raises further complications.Although modern readers know Pindar almost entirely as a composer of epinikia, otheraspects of Pindar's oeuvre, collected and organized in Alexandrian editions, were far

30 Harrison 1995b summarizes different views, and reaffirms Biicheler's suggestion (dating to 1889)that maiore plectro refers to panegyric in epic verse, a genre in which 1ulluscomposed (his twelvebook Diomedeia was highly praised).

31 Cf. Fraenkel 1957, p. 439: 'No Pindaric ode now, no elaborate poetry at all: the medium throughwhich he voices his enthusiasm is the homely versus quadratus'. Indeed, Horace inscribes thetrochaic chant 0 sol pulcher, 0 laudande sc. qui subegit Raetiam (vel sim.) into his ode; but the'homely' metre may disguise a more serious connection. The Latin septenarius is equivalent to theGreek trochaic dimeter cata1ectic, a metre used by Archilochus, arguably with generic significance,in his dithyrambic fr. 120 (see D' Angour, forthcoming); and Horace will have been aware of thesupposed etymological bearing of Greek dithurambos on Latin triumphus (Cairns 1972, p. 96). Ifthe reiterated (non semel) 'god of Triumph' is none other than 'twice-born' Dithyrambos (Bacchus)in one of his many guises, Odes 3.25 shows that the god is more than a 'homely' source ofinspiration to Horace: quo me, Bacche, rapis tui / plenum? ... nil parvum aut humili modo, / nilmortale loquar; dulce periculum est, /0 Lenaee, sequi deum ...

32 Odes 4.2.49-52.

33 Cf Fowler 1995.


better known to Horace and his contemporaries than they are to US.34 Odes 4.2 surveys the

range of Pindaric compositions: lines 10-12 speak of dithyrambs, lines 13-16 embracehymns, paeans, and encomia; lines 17-20 speak of epinikia, songs that memorialize betterthan statues (potiore signis / munera), and lines 21-24 speak of dirges, which preserve thevirtues of the departed from oblivion (moresque / aureos educit in astra nigroque / invidetOrcoi/" As we can tell from the fragmentary works, non-epinician songs contained many ofthe elements of Pindaric style, tone, and structure which we associate with his epinikia. Thecharacteristic elements of Pindaric composition - priamels, abrupt transitions, mythicalexempla, gnomai, bold usages of words and syntax, and his inimitable, seeminglyunregulated metrical patterns (numeri) - will have spanned different generic domains. Sowill his melodies (modi), now lost to us but no doubt heard in some form by Horace.36

Pindar, says Horace, soars like a Dircean swan, while he himself is a Matine bee,toiling on a small scale to construct his carmlna." While there is genuine modesty in theface of Pindar's elemental genius, Horace has other reasons to draw attention to thecontrast in their styles. He will not seek to imitate Pindar, one might suppose, because hehas created a poetic style in Latin that makes him distinct and original. It would not havebeen impossible for Horace to attempt Latin composition in the kind of complex, variableverse-forms used by Pindar." But the marvellously adapted, neatly contained, elegantlyric stanzas Horace uses were of his own refashioning; and he would have had no wish orreason to belittle the styles he had carefully devised in his quest to tum Greek music intoItalian measures." By this stage of his life, Horace was a poetic star in his own right, not

34 Feeney 2009, p. 204 estimates that Horace would have had roughly four times as much Pindar aswe have.

35 Freis 1983.

36 For the argument that Greek melody in Pindar's time was related to the pitch profiles of Greekwords. see D'Angour 2006a, pp. 276-80; I have proposed a practical reconstruction of Pindaricmelos on these lines (0'Angour, forthcoming).

37 The m's of Matinus / more modoque nicely imitate the humming of a bee (as observed to me byStuart Lyons); the bee itself is a Pindaric image (e.g. P.IO.54f., and cf the oft-repeated story ofPindar's being visited by a bee as an infant in Paus. 9.23.2 etc.), but here located in an emphaticallyItalian setting.

38 Horace acknowledges the possibility of doing so when he speaks in Epist. 1.3.5-13 about Titius'daring, asking fidibusne Latinis / Thebanos aptare modos studet auspice Musa? Cf Ovid EP4.16.27-8, une / Pindaricaefidicen tu quoque, Rule. lyrae.

39 Regarding numeris lege solutis and immensus, Horace could not conceivably have supposed thatPindar did not create metrical verses, even in his audaces dithyrambos: here Horace is clearlyplaying on the association of metrical licence with the non-antistrophic dithyrambs associated withthe New Musicians of fifth-century Athens (0'Angour 2006, pp. 270-73). The image of super notasripas confirms the correct interpretation, viz. that Pindar's metres do not conform to regular epodic,stanzaic or strophic patterns from ode to ode, in obvious contrast to Archilochus and the Aeolicpoets; cf Steinmetz 1964, p. 10, Freis 1983, p. 31.


a mere imitator; it is not surprising that he chose to fly with his own, not borrowed,wings.t"

The imagery of Odes 4.2 hints at a further purpose underlying the poet's deliberate,but less than total, Pindarizing. Writing praise poems for Augustus, or indeed quasiPindaric epinikia about the triumph of Caesar and his relatives, was to some extent anartificial exercise, and potentially a dangerous one. The notion of artifice in a premechanical age may be indicated by the contrast between constructed materials andnatural forces. It may be natural for bees to produce honey (mel, j.lEAl), just as it is naturalfor poets such as Pindar and Horace to produce songs (j.lEAT]). But bees also produce wax,with which they build their intricate beehives; and a craftsman such as Daedalus may usebeeswax to construct wings." Such devices, however, are not a natural means of flight,and an Icarus who flies too close to the sun will find that his artfully constructed wingsfail to sustain him. To marry nature and teaching, ars and ingenium, is Horace's explicitpoetic ideal in the Ars Poetica and elsewhere.f The balance of nature and artifice is clearin relation to Pindar, whom Horace represents not so much as a poietes (literally a'maker', as poeta used of Iullus in Odes 4.2.34 also seems to stress) as a force of nature,monte decurrens velut amnis. In his relations with the Princeps, Horace needed tocombine poetry with tact, and obvious poetic artifice may have been felt to obtrude ongenuine sentiment. A closer emulation of Pindar might have got the poet into deep waterin more ways than one."

Considerations of poetic texture may afford further insight into Horace's care not tostrike the wrong note in praising Augustus. What guided Horace's metrical choice incomposing the praise poems 4.4 and 4.14 in Alcaics rather than Sapphics? I would arguethat after the Carmen Saeculare of 17 BC, which required unprecedented attention to therhythm and melody of Sapphics, Horace attached particularly personal associations to themetre which made it unsuitable in his ears to be the vehicle of political or princely eulogy.This conclusion emerges from an analysis of Horace's changing metrical practice in hisSapphics, and broader considerations relating to that phenomenon." The revision ofGreek metrical patterns, with the tendency to regularize Greek anceps (a syllable thatmight be either long or short) as a long syllable, was a large part of Horace's personal

40 Cf Odes 2.20.1-3: non usitata nee tenui ferar / penna biformis per liquidum aethera / vates . ...The subsequent allusion to Daedalus (13-14, iam Daedaleo notior Icaro / visam gementis litoraBosphori) createsa significant echo in Odes 4.2.2.

41 Cowan 1908, p. 16 refers to the myth of Daedalus, and writes (p. 52) 'Beeswax retains itsductilityand tenacityundergreaterrangesof temperature thanany mineral, plant,or insectwax'.

42 AP 295-301, cf Epist. 1.19.1-14; Brink 1971, pp. 512f.,Commager 1962,pp. 20-31.

43 Cf Fraenkel 1957, pp. 434f., who notes that Horace was concerned not to attemptpericulosaeplenum opus aleae (Odes 2.1.6) in relationto Augustus, and suggests that 'what he had cometo feelfor him, gratitude, admiration, and finally affection, was not fit to be clad in the purple of amagnificent artisticconvention' (p. 439).

44 Rossi 2009 argues from developments in Horace's Sapphics that the poet was 'a Greek Iyristwithoutmusic' - apart fromthis singleoccasion. Even if this supposition is correct(Lyons 2010,pp.9-10 calls it 'scarcely credible'), it misses the wider significance for Horatian Sapphic compositionof the one occasion on whichhis Sapphics were performed to music.


contribution to the creation of Greek lyric in Latin; and in the case of Sapphics, Horaceadapted the Greek rhythm so that what had been a fourth-position anceps became aninvariable long syllable." As a result, the normal Latin word-stresses of Horace'shendecasyllabic lines, combined with a regular caesura after the fifth syllable, might haveinclined some Roman readers to hear the great majority of lines an effective four-beatstructure, with the main ictus evenly distributed over the first, fourth, sixth (short), andtenth syllables as underlined below: 46

lam satis terrisInteger vitaePindarum quisquis

I nivis atque dirae .I scelerisque lllirus .Istudet aernulgri ...

This is not, however, how Sappho's own poetry can have been articulated: the fact that thefourth syllable is anceps in Greek suggests that it will not have borne the ictus(corresponding to a footfall or thesis) on its own." The Greek rhythm implied by thisvariable fourth syllable is a three-beat line, based around the rhythm of the centrallypositioned choriamb (- u u _).48 The three main stresses accordingly fall (at more or lessevenly-spaced intervals) on the first, fifth, and tenth syllables, all of which are long."

<Qgivnul IJ-OL !illvos 'lCJos 8EQ1.CJLV

45 Catullus, by contrast, freely employs the same metrical licences as Greek models in his twoSapphic Odes, II and 51. Horace may have sought to avoid a trochaic opening rhythm (- u - u dumdi dum di) in favour of making it sound more like Pindaric dactylo-epitrites: with a regularly longfourth beat one effectively hears an opening epitrite (- u - -dum di dum dum) followed by a dactylicelement (- u u -) as one might in a passage of dactylo-epitrites, Cf Wilkinson 1940; Whitaker 1956.

46 A mnemonic I devised to illustrate this accentual pattern (stressed syllables underlined) runs:Conquering &JJlpho 's not an easy business: / Masculine fgdies cherish indeJl§!1dence; / Only goodmusic Jl§!1etrates the souls of / Lesbian artists. An evenly-spaced four-beat line of this kind wouldrequire time-adjustments, such as a slight lengthening of the fifth syllable before the caesura, andcreate what Wilkinson (1940, p. 133) characterizes as the 'barbarous Roman time' of a rhythmicallyinsensitive schoolboy. (Since the general pattern demonstrates my point, I pass over other possibleand secondary stressed positions in lines 1-3 of the Sapphic stanza).

47 However, a short beat in combination with a following long might together have been consideredthe locus ofa thesis. In the dochmiacs of the Rainer papyrus G2315 (Euripides Orestes 338-44), thepointing (stigmai) above the first and third elements of .:,.:': u - indicates upbeats, apparently

implying two theseis, one on the second position, the other on the fourth and fifth together, asbracketed in .:, i-r- [u-]: West 1982, pp. 114-15. The traditional English three-beat mnemonic'The wise kangaroos' should therefore be substituted in this case with a two-beat one, such as thatsuggested by Gail Trimble, 'That 01' man river': D'Angour 2006b, p. 492.

48 A long syllable in the fourth position in Greek Sapphics may have represented in practice a'dragged' syllable rather than a true long.

49 An illustrative mnemonic, with stressed syllables underlined, might run: l!1dependent metre isoverrated: / What's the use if nobody knows the dance-form?/ Wisely, Sappho chose to adopt astately / regular stanza.


This 'Greek' rhythm is firmly asserted by Horace in the first lines of Odes 1.10 and1.12 with their delayed caesurae.'" Mercuri facunde nepos Atlantis and quem virum autheroa lyra vel acri (it would be impossible for the e of facunde or a of heroa to bear astress). The deliberate use of this metrical pattern at the start of these Odes suggests thatHorace was signifying that, despite the apparent counter-indications of natural wordstress, his Sapphics too should be articulated with a three-beat line.51 The resulting lack ofcoincidence (in the great majority of verses) of primary ictus with the natural stress ofLatin words served to create both a contrapuntal texture within the line (similar to thatfound in Virgilian hexameters) and a satisfyingly varied contrast with the final adoneancoda (- v v - - dum-di-di dum-dum) where (as in Virgil) ictus and accent usually coincideto provide felicitous rhythmical closure. It seems likely that this was the kind of artisticeffect that Horace intended in creating the characteristic (but not invariant) verbalarticulation of his Sapphic verses.52

The incidence of delayed caesurae increases substantially in the Carmen Saeculare,and subsequently in the Sapphic poems of Odes 4.53 These facts must be related, as theygenerally have been, to the unique circumstances of the Carmen's realization; but theconsequences of that 'performance event' for Horace's subsequent compositional practicehave not been fully appreciated. Since the song was composed to be publicly performedby a chorus, Horace needed to insist on its correct rhythm and melody. He therefore tookpains to emphasize the correct ('Greek') rhythm to which it was to be sung andmarched. 54 With one foot (say the left) leading in each three-beat verse, the first two lineswould each have required six alternating footfalls (the sixth step falling in the pausebetween lines); while for the last couplet, the adonean coda adjoined seamlessly to thepreceding line would have created a concluding colon of four principal beats (seven

50 Also Odes 1.30.1. The use of delayed caesurae in these Odes' first lines is salient, and should notbe submerged in the statistics that Rossi (2009, p. 369) gives for the first book of Odes: he countsonly 7 'violations' (late caesurae) in 447 Sapphic hendecasyllables. In any case, the predominant'strong' caesura is, though standard Horatian practice, not a rule: 'violation' is a tendentious term,especially since Odes 1.10 on its own has three instances oflate caesurae in 15 hendecasyllables i.e.an incidence of20%! - essentially the same as that of the fourth book of Odes (see n. 53 below).

51 Cf Wilkinson 1963, p. 109. Ancient scholastic evidence points to precisely this articulation ofLatin Sapphics in the practice of well-schooled readers. Marius Victorinus, a fourth-centurygrammarian and rhetorician, indicates (De metris Horatianis 6.162 Keil) that the Sapphic line wastaught as articulated with three stresses (or five, if the gaps between the three principal stresses arefurther subdivided), but never with four: Becker 2010, pp. 162-3.

52 So also Wilkinson 1940 and cf Wilkinson 1963, pp. 106f.

53 The statistics are starkly informative. Rossi 2009, pp. 369f. gives figures as follows: the Sapphicsof Odes 1-3 have delayed caesurae in only 1.57% of hendecasyllabic lines, whereas the Carmen has33.34% and Odes 4 has 20.96%.

54 Singing or recitation does not require a strict regular ictus; marching in time does. Fraenkel1957,p. 378 dismisses Mommsen's conclusion that the Carmen was performed as a processional song('No compromise should be considered'), adding the scathing footnote (n. 5) 'It may, of course, besafely prophesied that the procession song will from time to time find a new champion'. Indeed, ithas found in Schmidt 2009 a most convincing one.


alternating footfalls, with an eighth, or more, to come in the gap between stanzas). 'Keepto the Lesbian foot and my thumb's controlling beat', Horace recalls telling his performersin Odes 4.6 (lines 35_6).55 His verses repeatedly demonstrate this metrical point byasserting the rhythmical ictus that is required when the caesura falls after the sixthsyllable:56

Deliae tutela deae, fuggcislyncas et cervos cohibentis arcu,Lesbium servate pedem meique

pollicis ictum.

The Sapphic Odes of Book 4 continue the new tendency to place the caesura late muchmore often than before, and the reason is now evident: until the Carmen, Horace was notthinking of his Odes as music - or at least, not music that required to be sung to a steadybeat.57 After the Carmen, he will have found it hard, when composing in Sapphics, not tohear, if only in his mind's ear, the sound of the music - that is, both the reinforced rhythmand the melody - that characterized the strikingly successful public performance of hispiece.58 Equally, he may have expected that readers and hearers of his Sapphic verses(particularly those in the educated elite circle around Augustus) would be bound to make asimilar aural connection to the song that had been so strikingly performed at the LudiSaeculares. 59 In the context of praise for the Princeps or other noble honorands, anysubsequent encomium composed in this metre would, therefore, potentially raise an

55 I doubt that pol/icis ictus can refer to the snapping of finger against palm to mark time (whichwould be digiti ictus); it is more likely to indicate the sweep of the thumb (with or without plectrum)across the strings of a cithara. Rossi's correct insistence that Odes 4.6 is a mimesis of song,therefore not itself sung, overlooks the way Horace's mimetic lines are meant to indicate preciselyhow the Carmen had been rhythmicized in performance; the poet is here 'singing along' to thatparticular Sapphic beat, and inviting readers to do so.

56 Cf Odes 4.11.34, where Horace advises to Phyllis to 'learn the refrain' - {emina, condiscemodos... Colin Sydenham has observed to me that the late caesura is only found in the last twostanzas of the ode, in which Horace gives, as it were propria voce (and accordingly proprio metro),his officious advice about Telephus.

57 This is not to deny that the Odes could have been set to music and sung in some cases and on someoccasions, as they were in later ages (cf Wille 1967, pp. 253f£). Rossi 2009, p. 362 overstates the casein claiming that 'the culture of Horace's time lacked the institutions to support, in both place andfunction, the musical performance of his poetry' (contra, Lyons 2010, pp. 43f.). Moreover, even if'editorial arrangement clearly reveals the bookish nature of the odes' (Rossi 2009, p. 362), they werenonetheless individually composed on different occasions over an extended period, and onlyretrospectively arranged in books.

58 Horace signals his delight at his success with the words (Odes 4.3.13-15) Romae principis urbium/ dignatur subo/es inter amabi/is / vatum ponere me charas, 'the offspring of Rome, prince of cities,sees fit to place me among the much-loved choirs of bards' .

59 In Odes 4.6.43-4 Horace imagines one of the girls who performed his Secular Song, now married,reminiscing reddidi carmen doci/is modorum / vatis Horati. It is significant that the only use of hisname in the Odes occurs in relation to the Carmen Saecu/are; cf Fraenkel 1957, pp. 406-07 on thissphragis.


embarrassing spectre: CARMEN COMPOSUIT Q. HORAnus FLACCUS. 6o It was upto Augustus, not the poet, to call the tune: Horace could not appear to be indulging inundue self-advertisement by immodestly 'harping on' his success. Accordingly, whileHorace's recusatio of (and in) the Pindaric style, 4.2, and his highly personal invocationto the god of poetry, 4.6. are composed in the Sapphic metre he had once chosen foroffering signal praise to the Princeps (1.2 and 1.12) and subsequently for the CarmenSaeculare, in Odes 4.4 and 4.14 he reverts to the Alcaic metre of the grand, sacerdotal,Roman Odes of the third book.?' Horace was not Pindar, nor was meant to be.

60 For the inscription see Fraenkel1957, pp. 367f.

61 The only other Ode in book 4 in Alcaics is 4.9, with its unprecedented direct praise: est animustibi I rerumque prudens et secundis I temporibus dubiisque rectus (34-6) etc. The associations of themetre at this stage lend decisive weight to the standard view that, pace Ambrose 1965, this is aserious eulogy of Lollius.

Horace's Victory Odes

Documents